Las Cruces, New Mexico | February 10, 1990
The killers tried to leave a silent room. They shot seven people, lit papers on a desk, and walked out of a bowling alley with less than five thousand dollars. But at about 8:30 that Saturday morning, a twelve-year-old girl reached a corded telephone. Her call carried two things through the smoke: the urgent fact that people were still alive, and a time-stamped boundary around a crime that has remained unsolved for more than thirty-five years.
I began this reconstruction with the building rather than with a suspect list. A bowling alley is a collection of very different rooms. The public lanes are broad, bright, and meant to be seen. The manager's office is small, controlled, and connected to money. The telephone opens a route outside the building. The fire creates a destructive layer over everything beneath it. Treating those spaces separately makes the case less sensational and more precise.
The quiet hour before the lanes opened
Las Cruces Bowl stood at 1201 East Amador Avenue. The City of Las Cruces now describes it as a popular local bowling alley. Saturday, February 10, 1990, began before the doors would fill with customers. Contemporary reporting said a youth league was due later that morning; one league official told UPI that dozens of children might have arrived only about half an hour after the attack. That fact does not prove the gunmen studied the league schedule, but it defines how narrow the interval was.
Inside were three adults and four children. Stephanie Senac, thirty-four, managed the alley. Her daughter, twelve-year-old Melissia Repass, was with her, as was Melissia's thirteen-year-old friend Amy Houser. Ida Holguin, thirty-three, worked as the cook. Steven Teran, twenty-six, was the alley mechanic. His daughters, six-year-old Paula Holguin and two-year-old Valerie Teran, were with him. The shared surname of Ida and Paula has caused occasional confusion; the city's current account explicitly states that Ida was not related to Paula or the Terans.
The public record gives a simple reason for the early presence of money: the business was preparing its deposits. KVIA reported that the manager and others were in or brought into the office while two men took thousands of dollars from the safe. The city's 2025 release describes the amount as undetermined but believed to be less than $5,000, made up of cash and personal checks. That combination matters. Checks are traceable and awkward to spend. Their theft may indicate speed, indiscriminate grabbing, or incomplete knowledge of what the safe contained.
It is tempting to write this morning as an inevitable approach toward violence. It was not inevitable to the people inside. They were doing ordinary work in a familiar place. The documentary facts only become ominous in retrospect: a quiet hour, an office with the previous day's receipts, doors not yet busy with league families, and two men who could control a small group before the public area came alive.
Seven people, not seven pieces of evidence
A crime summary compresses human beings into ages and injuries. That compression is useful for a timeline but harmful if it becomes the entire story. Steven was an employee and a father who arrived with his daughters. Amy was a teenager spending the morning with a friend. Melissia was a child in her mother's workplace. Ida and Stephanie were beginning a workday. Paula and Valerie were six and two. Their relationships explain why several children were in a commercial office before opening; they should not be reduced to a puzzle about offender behavior.
Bowling-alley mechanic and father of Paula and Valerie. He died at the scene.
Steven's older daughter. She died at the scene.
Steven's younger daughter. She was taken to a hospital and died shortly after arrival.
Melissia's friend. She died at the scene.
Manager and Melissia's mother. She survived the morning but died in 1999 from complications attributed by the city to her injuries.
Stephanie's daughter. Wounded, she reached the phone and called 911.
Bowling-alley cook. She survived and later spoke publicly about the attack and its lasting effects.
The current city account says first responders found Steven, Paula, and Amy dead. Valerie was rushed to a hospital but died soon afterward. Melissia, Stephanie, and Ida were hospitalized and survived the initial attack. Stephanie's death nine years later is part of the event's human cost even though public summaries commonly refer to four deaths that day. This distinction is not arithmetic trivia; it preserves the difference between the immediate scene and the injuries that continued to shape a life.
Early news reports are not perfectly consistent with later official information. A UPI dispatch published two days after the murders listed Steven as thirty and called the two-year-old child Victoria; the city's current release identifies him as twenty-six and the child as Valerie. I use the later Las Cruces account for names and ages while retaining the early report for what officials said at the time. A responsible cold-case narrative does not silently combine contradictory editions of the record.
The office: where a robbery became something else
According to the City of Las Cruces, two gunmen forced all seven people into an office. KVIA's 2025 account adds a sequence: Stephanie, Melissia, Amy, and Ida were already under control while the men looted the safe; Steven and his daughters then encountered the event. The attackers shot everyone at close range. Reporting identifies a .22-caliber pistol, though the complete ballistic file has not been made public.
The difference between entering an ongoing robbery and being present from its first second may explain the changing number of people under control, but it does not explain the decision to shoot. Robbery supplies an immediate economic motive. It does not, by itself, supply the motive for attempted witness elimination. That second decision must be analyzed separately. Did one or both men believe they had been recognized? Did they arrive intending to leave no witnesses? Did panic, an unexpected arrival, or a prior plan change the offense? The evidence released publicly does not choose among those possibilities.
| Publicly supported point | What it permits | What it does not prove |
|---|---|---|
| Two men controlled the office. | Coordination between at least two offenders. | Their relationship, roles, or whether anyone else assisted outside. |
| Money and checks were taken from the business. | A real financial objective, not merely a staged appearance. | That money was the only motive or that the amount was known in advance. |
| Seven people were shot. | An intentional attempt to prevent resistance, identification, or both. | Which offender fired each shot or whether both fired. |
| Papers were set on fire. | Deliberate destruction or concealment is the most direct explanation. | That the offenders understood exactly which forensic traces fire would destroy. |
| The youth league had not yet arrived. | The attack occurred within a low-occupancy window. | That the offenders possessed the league schedule. |
The office was both a money room and a confinement room. That dual use shapes the strongest hypothesis: the men selected a business at a time when cash would be accessible and witnesses few. The arrival of Steven and his daughters may have enlarged a situation they expected to contain. Yet the execution-style shooting and the fire show more than opportunistic theft. They show an effort to close every route by which the room could speak afterward.
The call that broke the intended silence
At about 8:30 a.m., Melissia reached the telephone despite multiple gunshot wounds and called 911. Public broadcasts have replayed portions of the recording for decades. I will not transcribe it in full. Its investigative value can be described without turning a child's pain into spectacle.
First, the call establishes that at least one victim was conscious after the gunmen left. Second, it places the fire and the shooting within the same immediate episode. Third, it triggered simultaneous police, fire, and medical response. Fourth, it counted seven injured people before rescuers entered. The caller's statements were not a later reconstruction shaped by years of publicity; they were contemporaneous observations made under extreme trauma. That makes the recording exceptionally important, but not infallible in every detail.
The National Institute of Justice's eyewitness-evidence guide treats the emergency call as the first stage of witness information. A dispatcher must obtain lifesaving facts without turning that moment into a full investigative interview. Memory recorded close to an event avoids some later contamination, while overwhelming stress, injury, attention to a weapon, smoke, and fear can restrict what a witness perceives. Both propositions can be true. Respect for Melissia's courage does not require pretending that trauma creates a perfect camera.
In this case the call also changed the physical scene. Firefighters were not responding to an empty burned office hours later; they were responding to a reported active fire with wounded people inside. Rescue properly came before laboratory preservation. Any later criticism of the scene must begin with that priority. A living child on the phone defines the first obligation.
The fire layer
The gunmen set fire to material in the office before leaving. The UPI report from February 12 said papers on a desk had been ignited and firefighters extinguished the blaze quickly. Later accounts describe the fire as an apparent attempt to destroy evidence. That inference is strong because the fire began in the room containing the victims, money, telephone, and likely offender contact surfaces. It was not a separate accidental event elsewhere in the building.
Fire does not simply make evidence disappear. It transforms the scene unevenly. Heat can damage biological material and latent prints. Smoke and soot can coat surfaces. Water can move ash, dilute residues, and carry small objects. Rescue and suppression require people to cross the room, open paths, move obstacles, and treat victims. Some materials survive in protected gaps, under furniture, inside containers, or in areas where heat and water were limited. The result is not a blank room; it is a layered room whose original arrangement is harder to read.
NIJ guidance for fire and arson scenes emphasizes documentation, controlled access, careful collection, contamination prevention, and a complete chain of custody. That guide was published years after this crime, so it should not be used to score the 1990 response against a later checklist. It is useful for explaining the technical problem: fire suppression, salvage, movement, and uncontrolled entry can all alter trace evidence, even when every action is taken for a legitimate reason.
Here the assailants appear to have made a destructive choice, but they failed at total erasure. They did not prevent the 911 call. They did not prevent witness survival. They did not burn the entire building. They left behind a physical file substantial enough for later investigators to seek fingerprints, ballistics, and DNA. What the fire did was reduce quality and context, turning potential identifications into partial comparisons and uncertain samples.
What a burned scene can still say
Popular retellings often make one of two errors. The first claims the fire destroyed everything, which is contradicted by the continuing forensic work. The second claims modern technology can inevitably recover a profile from whatever remains. The 2025 reporting rejects both extremes. Las Cruces Police told KVIA that investigators had recently explored testing with an outside company capable of working with very small amounts of DNA, but the available evidence did not yield a usable result.
A failed DNA attempt is information about the sample, not proof that no offender material ever existed. Biological material may have been consumed, diluted, mixed, collected on an unsuitable substrate, exhausted in prior testing, or present below a laboratory's validated threshold. It may also be absent from the surviving items. Modern sensitivity cannot recreate material that was never preserved.
Other evidence categories may still matter. A fired bullet can carry class and individual characteristics if it remains sufficiently intact. Cartridge cases, if present, may preserve marks from a weapon's mechanisms. Checks can create a paper trail if they were negotiated or discarded. Latent fingerprints can be compared repeatedly as databases grow, provided their quality supports a search. A suspect's geography, employment, incarceration history, vehicle access, and connection to similar robberies can be tested against the timeline even when DNA is silent.
NIST defines crime-scene work broadly as documenting, locating, preserving, and collecting material whether it is hidden or charred. The word that matters is preserving. A cold case cannot retest the room; it can only retest what entered evidence storage with enough provenance to remain interpretable. Every evidence box therefore has two histories: what the object was in the office, and what happened to it during thirty-five years of custody and examination.
Two descriptions, two composites, no names
In the first days, police circulated descriptions of two men seen leaving the area. UPI reported that the information came from neighbors and that investigators produced composite drawings. The preliminary descriptions were specific: one man was estimated at forty-five to fifty years old, about five feet six inches and 140 pounds; the other around thirty, about five feet ten inches and 170 pounds, with distinctive hair. Both were described as dark-complexioned. The city's current summary says investigators believe both men were Hispanic and spoke fluent English.
These labels must be handled carefully. A composite is not a photograph. Height and age estimates made at a distance can be wrong. Ethnic appearance is especially broad and subjective. Fluent English does not establish nationality, birthplace, residence, or immigration status. A resemblance to a drawing is a lead-generating observation, never a sufficient identification.
Time changes the value of the drawings in opposite directions. Age progression can help a witness imagine how a face might look decades later. But every added year also enlarges the number of people who resemble a generalized image. Repeated television exposure can make a composite feel familiar even to someone who never knew the offender. The strongest new tip would therefore contain verifiable facts beyond appearance: access to the alley, unexplained cash or checks, a contemporaneous admission, a matching weapon, knowledge of the fire, or a route across the border region that fits independent records.
The initial descriptions do not tell us which man led, which man carried the money, which man fired, or whether each participated in every act. They do establish a working minimum of two offenders and provide features investigators can compare. In a case with surviving witnesses, the ethical approach is to preserve the original descriptions and interview history, not to overwrite them with internet speculation.
Did the killers know the business?
The amount taken was modest compared with the violence. The men reached the office, accessed the safe, took cash and checks, controlled seven people, and attempted to burn the room. Investigators told UPI in 1990 that they believed the men had been at the bowling alley previously and knew where the money was. This is often compressed into the phrase "inside job." That phrase goes farther than the evidence.
Knowledge can come from many sources. A former employee, vendor, customer, league participant, maintenance worker, family acquaintance, or habitual observer could learn the building's rhythm. A person could simply visit during business hours and see where the office was. Repeated casing from the parking lot could reveal when employees arrived and when youth bowlers appeared. An accomplice could share information without entering. None of these possibilities should be attached to a named person without corroboration.
I tested three versions of the knowledge hypothesis:
| Model | What it explains | What it still needs |
|---|---|---|
| Observation from outside | The low-traffic window, employee arrival pattern, and easy departure before customers. | How the men knew the safe's location and whether a useful deposit would be present. |
| Ordinary prior access | Familiarity with the office, doors, and routine without requiring an insider accomplice. | A record or witness tying a particular pair of men to the business before February 10. |
| Information supplied by someone | Precise timing and money knowledge, if those were truly precise. | Evidence of communication, payment, relationship, or conduct after the crime. |
The first two models require fewer hidden actors. They are therefore preferable starting points. The third cannot be dismissed, but it should not become a rumor machine. The public file does not show that an employee or family member arranged the robbery. Familiarity with a commercial building is not the same as belonging to the circle that operated it.
Robbery, targeting, and the problem of excess violence
A planned commercial robbery is the most economical explanation for entry, safe access, and stolen money. The difficult question is why the offenders escalated to attempted mass murder. Several explanations can fit the same external facts. They may have planned from the beginning to kill witnesses. One may have been recognized. Steven's arrival with his children may have disrupted a simpler plan. The offenders may have believed the fire would make the deaths look different or eliminate identifying traces. A personal motive could have been hidden inside a real robbery.
Only the first of those propositions is necessary to explain the observable sequence: two men chose a lightly occupied business, sought money, controlled witnesses, used lethal violence, and tried to damage the scene. The personal-target theory adds a motive that public evidence does not require. It becomes stronger only if one victim was treated differently, if the offenders ignored accessible money to reach a person, if threats showed personal knowledge, or if independent records reveal a conflict. None of that has been publicly established.
The fact that some cash was reportedly left behind, mentioned in later CNN coverage, is not proof that robbery was staged. Offenders under time pressure frequently leave value. Checks may have been grabbed with bills. The growing number of victims, a fire that needed time to catch, and the approach of customers all created pressure to flee. A partially completed theft is compatible with a genuine robbery.
The violence also does not prove professional expertise. The attempt to eliminate witnesses failed. The fire was extinguished. Three people initially survived. Two offenders were seen in the area and described. These outcomes suggest determination and brutality, not necessarily sophistication. A disciplined analysis should not promote the killers into criminal masterminds simply because they remain unidentified.
The border is geography, not an answer
Las Cruces lies near major routes and within an hour's drive of El Paso and the international border. In 1990 police considered that the men could have fled toward Texas or Mexico, and checkpoints were reportedly established on routes out of the city. The absence of an arrest at those checkpoints tells us only that the offenders were not identified there. They could have left before the net was organized, used local streets, changed vehicles, remained nearby, or passed unnoticed.
Border geography expands the investigative map. It does not supply ethnicity, cartel involvement, or a foreign escape by itself. The current official description of the men as Hispanic and fluent English speakers is an eyewitness-based lead, not evidence that they crossed a national boundary. Las Cruces and El Paso are binational communities in daily practice; countless innocent people fit broad demographic descriptions and travel the same roads.
A useful geographic analysis would remain concrete. Which parking-lot exits were visible? What businesses opened before nine? Were fuel stations, banks, or traffic cameras present in 1990? Which employees or delivery drivers used East Amador Avenue that morning? Were the stolen personal checks ever found, deposited, or reported through financial channels? Which vehicles were seen, and when were those accounts first recorded? Public retellings rarely answer those questions, but the original investigative file may.
The location also matters for offender comfort. Selecting an early Saturday window suggests either reconnaissance or experience with similar businesses. Leaving without immediate capture suggests a planned first route, but not necessarily a long-distance one. The simplest route could have ended a few blocks away in another car, apartment, or workplace. Distance should be measured from evidence, not from the most dramatic theory.
Survivor memory is evidence with a duty of care
Ida Holguin later described being seized and told the event was a hold-up. Melissia spoke publicly about her decision to reach the telephone. Their accounts are indispensable because they place language, movement, and human behavior inside a room that the fire damaged. They are also the memories of people who endured catastrophic violence, hospitalization, grief, and decades of public attention.
The NIJ eyewitness guide recommends documenting individual recollections, the conditions under which they were obtained, and the confidence attached to specific details. That structure prevents a common error: treating a witness as either perfectly reliable or useless. A person may remember a command accurately, estimate height poorly, recognize a voice, miss an object outside the focus of attention, or gain clarity about one sequence while dates become less precise. Each component should be tested separately.
Repeated interviews create additional obligations. Investigators should know which features appeared in the first statement, which emerged after viewing composites or media, and whether witnesses discussed the men with one another. Public audiences usually see the final, polished account. A case file can preserve the evolution. That history is essential if a new suspect is ever presented in a lineup or if a confession is compared with facts never publicly released.
Survivors do not owe the public endless testimony. Their decision to speak, remain private, attend a vigil, or refuse another retelling has no bearing on the truth of the crime. The ethical purpose of reviewing their accounts is to locate testable information, not to demand emotional performance.
Why the case may still be solvable without a miracle sample
In 2025 Detective Kenny Davis told KVIA that even an outside laboratory using small-sample methods could not obtain the hoped-for DNA result. He also emphasized tips: something remembered from a friend or family member, even if it seems small. That is not a ceremonial request. A two-offender crime creates social leakage. Each man had the potential to tell a partner, relative, cellmate, spouse, or friend. Money, checks, weapons, burns, unusual travel, abrupt absence, or fear of news coverage may have been observed by someone else.
A useful tip should be judged by detail that can be verified and was not widely publicized. "He looked like the sketch" is weak. "Two relatives returned before noon, destroyed clothing, possessed checks from a bowling business, and used a particular vehicle" is stronger because records and witnesses can test it. A claim about a .22-caliber firearm gains value if ownership can be placed before the murders and ballistics survive. A confession gains value if it contains scene knowledge protected by investigators.
Database growth may still assist fingerprints or firearms evidence if suitable material exists. Digitized archives can make old employment, property, court, prison, and vehicle records searchable. Genealogy is useful only if a qualifying biological profile exists, which police currently say they have not obtained. The most realistic breakthrough may therefore be cumulative: a credible name from a tip, records that place that person in the right environment, and old physical evidence used for confirmation rather than discovery.
There is also a time limit that technology cannot reverse. Witnesses die, memories fade, informal records disappear, and families discard objects whose relevance they never knew. That makes the present request for information urgent without making it desperate. The case needs a verifiable fact, not another theory repeated loudly.
My reconstruction and verdict
I place the strongest explanation in four stages. First, two men selected Las Cruces Bowl during a predictable low-occupancy period when business receipts were accessible. Their selection probably reflected prior observation or ordinary familiarity with the building; a secret insider is not required. Second, they established control in the manager's office and took cash and checks. Third, the appearance or presence of additional witnesses increased the problem they believed they had to solve. Fourth, they attempted to eliminate witnesses and damage the room before using a prepared exit route.
This model fits the smallest number of confirmed facts without inventing a personal target. It also explains why the amount stolen can be small: the primary plan was a commercial robbery, and the later violence was about escaping identification. I cannot determine whether the killers entered already committed to murder or escalated after being seen. The difference would be legally and psychologically important, but the public evidence does not expose it.
I give less weight to elaborate cartel, drug-debt, or family-vendetta stories because they add hidden relationships without publicly documented support. Proximity to Mexico is geography. Severe violence is behavior. Neither is proof of an organization. The men may have been experienced robbers, but the failed fire, surviving witnesses, and public descriptions show that the crime did not unfold with perfect control.
The decisive evidence gap is not one missing fingerprint. It is the absence of a publicly known bridge between the office and a named pair of men. The 911 call fixes the end of the event. Survivors describe the interior. Witnesses describe two men outside. The safe and burned papers establish purpose and concealment. What remains missing is a trace, record, weapon, check, vehicle, or private admission that connects those layers to identities.
That is why the title of this investigation is not about an unsolved riddle. It is about a call the fire could not erase. The killers' final act was designed to remove evidence. Melissia's action restored time, location, victims, and urgency to the record. It saved the possibility of survival and made the silence incomplete.
What would change the conclusion
I would revise this assessment if original records showed that the men demanded a specific person, ignored the safe until after the shootings, asked for a particular document, or used knowledge unavailable through ordinary observation. I would also revise it if ballistics tied the firearm to a known offender, stolen checks appeared in another case, or a protected-detail confession independently matched the office.
Until then, the correct conclusion is bounded: two men committed a planned robbery and a deliberate attempt to silence seven witnesses; they likely had some knowledge of the bowling alley's routine; the fire reduced but did not eliminate the evidence; surviving descriptions remain leads rather than identifications; and no publicly supported theory names the killers.
The Las Cruces-Dona Ana County Crime Stoppers reward announced by the city in 2025 is $25,000 for information that helps identify the men responsible. Tips may be submitted anonymously at 1-800-222-8477, through New Mexico Crime Stoppers, or through the P3 Tips system. Anyone with direct knowledge should contact investigators rather than publish a name online.
Sources and research notes
- City of Las Cruces, "Las Cruces Mass-Shooting Unsolved After 35 Years" (February 5, 2025) - current official summary, victims, time, location, amount, fire, offender description, reward, and tip channels.
- UPI Archives, "Laboratory experts search for clues in bowling alley" (February 12, 1990) - contemporaneous police statements, fire location, early descriptions, youth-league timing, and initial investigative posture. Early name and age discrepancies are not adopted over the current city account.
- KVIA, "Las Cruces bowling alley massacre unsolved after 35 years" (February 10, 2025) - event sequence, reported weapon caliber, current detective comments, and unsuccessful recent DNA testing.
- KVIA, "Victim recalls horrific shooting on 32nd anniversary" (February 10, 2022) - reward history and renewed forensic review.
- KVIA, "Survivors, Filmmaker Discuss Bowling Alley 'Nightmare'" (February 2, 2010) - Ida Holguin and family perspectives.
- KVIA, "Las Cruces bowling alley massacre unresolved after 23 years" (January 14, 2014) - survivor account and anniversary reporting.
- KOAT, "Family of Las Cruces Bowl Massacre continue to search for answers" (February 8, 2023) - victim-centered anniversary account and the emergency call's place in the case.
- CNN transcript, Anderson Cooper 360 Degrees (January 20, 2011) - historical case reporting, escape checkpoints, surviving witness interviews, and the distinction between published theories and evidence.
- National Institute of Justice, "Fire and Arson Scene Evidence" (2000) - preservation, documentation, contamination, and chain-of-custody principles.
- National Institute of Justice, "Eyewitness Evidence: A Guide for Law Enforcement" (1999) - emergency-call, interview, composite, and identification procedures.
- National Institute of Justice, "Crime Scene Investigation: Guides for Law Enforcement" (2017) - systematic scene preservation and integration of physical evidence with witness information.
- National Institute of Standards and Technology, "Forensic Science - Crime Scene" - modern overview of documenting and recovering evidence, including charred material, firearms evidence, and biological preservation.
Editorial method: Confirmed facts are attributed to the current City of Las Cruces release and identified reporting. The reconstruction sections are the author's inference and are labeled as such. The ten illustrations are newly generated editorial scenes; none depicts an actual victim, suspect, or verified crime-scene arrangement.