Dennis Dechaine, shown here in 2012, was convicted of the 1988 murder of 12-year-old Sarah Cherry. He has long maintained his innocence. Portland Press Herald photo by John Ewing

BOWDOIN — Thirty-four years ago this summer, 12-year-old Sarah Cherry was abducted in broad daylight from a home where she was babysitting. Her body was found two days later mostly hidden under leaves and brush in the woods about 3 miles away. She had been tortured with a knife, sexually assaulted with sticks and strangled to death.

A new round of DNA testing of crime scene evidence could help finally resolve a question that has haunted Maine ever since: Did Dennis Dechaine really kill her?

Some think the real perpetrator might be Richard Marc Evonitz, a notorious serial killer of children who was in Maine at the time and whose modus operandi fits closely with the horrific events that occurred in the wooded backroads of Bowdoin on July 6 and 7, 1988.

“The fact was that this guy was in the area, his fiancée isn’t with him, and he knew how to hotwire cars,” said Deirdre Enright, the University of Virginia School of Law professor who founded that state’s Innocence Project and freed a man wrongly convicted for the 1996 murder of two women with Maine ties after genetic tests showed Evonitz likely committed them. “I think not looking at him would be foolish, especially when you compare the likelihood of him doing this compared to Dennis Dechaine doing this.”

Dennis J. Dechaine, a 31-year-old farmer from nearby Bowdoinham, was convicted of Cherry’s murder and is serving a life sentence at the Maine State Prison in Warren, where he has been imprisoned for 33 years. Despite robust circumstantial evidence linking him to the crime scene, there have long been doubts about his culpability. The most difficult to dismiss: forensic science conclusively indicating that the girl was alive at least six hours after Dechaine – who had been injecting himself with speed in the woods and claimed to have become lost – had been picked up by police.

Though the facts about Cherry’s time of death were discernible at Dechaine’s 1989 trial, they were not recognized by his defense team at the time, nor during his unsuccessful 1989 appeal and his 1992 motion for a new trial. By the time his defenders realized their importance, it was procedurally too late: Courts dismissed them from consideration because, in the words of a 2015 court decision, “Maine has never recognized a freestanding claim of actual innocence as grounds for post-conviction relief.”


Dechaine has what may be his final chance at bat, however. Maine law does allow for the possibility of DNA tests to be conducted and exculpatory results potentially to be presented to the courts. On July 22, Knox County Superior Court judge Bruce Mallonee granted Dechaine’s request to test crime scene items using new DNA collection techniques that have been shown to collect several times more material from many surfaces as conventional swabs.

“We’re pleased with the ruling to allow the retesting,” said Dechaine’s pro bono attorney, John Nale, who normally specializes in elder law. “Where we go from there is just working in agreement with the attorney general’s office on who is doing the retesting.”

A police photo of Dennis Dechaine taken on July 6, 1988, the day Sarah Cherry disappeared. Shawn Patrick Ouellette/Staff Photographer

Asked if it was his intention to try to match any genetic profile that might be recovered from the objects found on Cherry’s body against potential perpetrators other than Dechaine, Nale responded emphatically in the negative. “No, no, no, nope,” he said. “That doesn’t come into the picture right now. The idea is to see if the DNA can be collected and if it can be tested and if a profile can be determined and if any of it is hopefully readable and can be compared to Dennis’.”

Experts on post-conviction exonerations said an absence of readable Dechaine DNA is unlikely to be enough to make a difference. Rather, his team would have to be able to locate crime scene DNA of someone else who could be convincingly shown to have been the killer. The standard is that the new evidence, placed alongside the evidence in the existing case record, would make it “probable” that a jury would find him innocent.

“That’s a difficult showing to make,” said York attorney Rory McNamara, who specializes in post-trial criminal litigation. “There is no guarantee that the murderer of Sarah Cherry – whoever it was – left DNA on her clothing or other items up for new testing. Likewise, there is no guarantee that the source of any DNA found on her body came from the murderer.”

“The upshot is, even if the results come back favorable to Dechaine, his lawyers will probably have a lot of work to do to win a new trial,” he said.


Alternative suspects abound in the case. Some were on Dechaine’s defense team’s radar decades ago. Their 1992 bid for a new trial was built around one of them, Douglas Senecal, who at the time of Cherry’s murder was facing charges of having had unlawful sexual contact with a 13-year-old girl who was friends with Cherry. (The court ruled their case for Senecal having killed Cherry “is no more than speculation and conjecture.”)

There was also the man boarding with Dechaine and his wife in Bowdoinham who had access to Dechaine’s truck, lacked an alibi for the day of the crime, and refused to let police search or photograph his room. There was the husband in the couple who hired Cherry to babysit that day whom a local waitress found creepy and who, asked about the murder years later, allegedly told a private investigator “The Lord will forgive anything.”

There was a man living in a trailer down the road from where Cherry was abducted who would soon be convicted of two rapes and gross sexual assault in the two years prior to the murder – including 11- and 12-year-old victims. The state police detective who was then investigating the later crime, Alfred Hendsbee, saw two sets of barefoot prints – one child-size – in the man’s driveway while responding to Cherry’s abduction but inexplicably decided not to investigate further. This man was arrested again in 2005 for having abducted a 14-year old girl.

Other suspects are not as well-known. David Guffey, a registered sex offender who lived on the road where Cherry was abducted, has been suspected to be Cherry’s killer by several people in town including one of his victims, who was Cherry’s classmate.

Then there’s a man who nobody involved in the Cherry case even knew existed until 14 years after her murder: Richard Marc Evonitz.



Serial killer Richard Evonitz

Unlike the other suspects, Evonitz is dead. He shot himself in the head after police surrounded his car on June 27, 2002. They’d been pursuing him for days since his last victim, 15-year-old Kara Robinson, managed to escape while he was sleeping after raping her for 18 hours in his apartment in Columbia, South Carolina. While police were closing in, he called his sister and told her he had “committed more crimes than I can remember.”

Detectives found this was likely true. In his apartment they found – in addition to the pink, wire-reinforced cuffs he had restrained Robinson with – nude photos of young girls, hundreds of pornographic images of 9- and 10-year-olds on his computer, and hundreds of pornographic videotapes including one of him molesting a young girl and another of him masturbating to Polaroid photos of other children. There were a large number of girls’ panties. And there were newspaper clippings and his own handwritten notes about the unsolved 1997 abduction and murders of Kristin and Kati Lisk, sisters ages 15 and 11, in Spotsylvania County, Virginia.

Investigators quickly found that fibers from Evonitz’s car trunk, a blanket in his apartment, the fuzzy pair of handcuffs he’d bound Robinson with, and the carpet of the Fredericksburg, Virginia home he’d lived in in the 1990s matched those found not only on the Lisk sisters but also on the body of 16-year-old Sofia Silva, who had been abducted, raped, murdered and dumped in a shallow pond in King George County, Virginia, in September 1996. His hair had also been found on all three bodies, and Kristin Lisk’s hand and palm prints were found on the inside of his car’s trunk.

Evonitz followed a pattern. All four girls – Silva, the Lisks and Robinson – were abducted from their semi-rural homes in broad daylight without any sign of a struggle. In Robinson’s case Evonitz had approached her nonchalantly pretending to be a magazine salesman, confirmed she was alone, pulled a gun and forced her to climb into his car trunk with a gag balled up in her mouth. He drove rural back roads scoping out potential victims but, as in Robinson’s case, was willing to switch targets if his intended victim didn’t follow her pattern. Once at home – he planned his abductions when his wives or girlfriends were away or had left him – he bound his victims, subjected them to long and repeated sex acts, and strangled them to death.

Federal investigators believed he had killed others. Evonitz had a 1987 conviction for masturbating in front of a 15-year-old girl in Jacksonville, Florida, but a clean record thereafter. Criminologists say there are rarely big gaps in serial killers’ careers.

“What really nags at us is he committed the three murders in Spotsylvania in 1996 and 1997, and then you jump ahead to the abduction in South Carolina,” FBI agent Donald Thompson told The Washington Post in 2002, shortly after Evonitz’s role in the Silva and Lisk murders was solved. “What was he doing? Did he commit more acts? With these kind of people, there are certain urges, perversions that drive them.”


A federal task force was set up to parse Evonitz’s life and to link him to other unsolved crimes in places he spent time in, including an eight-year stint in the Navy, where he served as a sonar technician aboard a frigate, the USS Koelsch.

Evonitz and the Koelsch were based in Portland from May 8, 1988, to May 31, 1989, while the ship was undergoing a refit at the Bath Iron Works facility then located near the present site of the Ocean Gateway cruise ship terminal. Cherry was murdered two months into this period. But when federal investigators reached out to the Maine State Police and Portland Police Department in 2002, they were told there were no unsolved cases that matched the profile.

The task force asked local law enforcement across the country for any information they had on Evonitz’s movements, but both the state and Portland police told the Portland Press Herald/Maine Sunday Telegram they have no records relating to this request. The task force never definitively tied Evonitz to any other crimes and has since disbanded.


Deirdre Enright, the founder of the University of Virginia Law School’s Innocence Project, has provisionally tied Evonitz to another double murder, leading to the release of Darrell Rice, who had been falsely convicted for the crime.

The victims – Julianne Williams and Laura Winans – had met while at Maine’s Unity College and were bound, raped and murdered while camping in Virginia’s Shenandoah Valley in May 1996. Enright and her team of law students exonerated Rice by securing DNA tests that excluded him as the perpetrator. The sample was consistent with Evonitz’s DNA profile, which was taken by the FBI after his death but is not available in the FBI database typically used by law enforcement to find matches. (For privacy reasons, federal law carefully limits which profiles can be included in the database; because he couldn’t be posthumously charged for his murders, Evonitz does not meet the criteria.)


Enright and her law students have taken up where the federal task force left off and are sifting through cold cases Evonitz may have committed. She says the parallels between the Cherry case and Evonitz’s other murders are significant.

“These murders required a perpetrator experienced in stalking, planning and controlling; they were no one’s first,” she said. “Evonitz – a known serial killer, in the same area, alone – makes a far better suspect than Dennis Dechaine.”

Like the other girls, Cherry was abducted from a semi-rural home in the middle of the day without any sign of a struggle. She was bound with sophisticated knots, gagged with clothing stuffed in her mouth, sexually assaulted and asphyxiated. She had been terrorized over a period of many hours before being killed. Her killer kept her panties and partially redressed her body.

Evonitz committed his murders when he was either single or when his wife or girlfriend were away on trips. In July 1988, 25-year-old Evonitz was living alone in Portland aboard the Koelsch, having left his 16-year-old fiancée behind in South Carolina. He had a car – a white Toyota Corolla – and shopped for groceries exclusively at the Brunswick Naval Air Station commissary, 13 miles south of Bowdoin, the fiancée, Bonnie Gower, told the Press Herald. While he had duty shifts on the ship, he had little to do when off duty, and ample free time.

Five weeks after Cherry was murdered he married Gower and rented a ramshackle basement apartment at 240 State Street in Portland. Gower joined him in Maine in October, she said.

Gower – who divorced Evonitz in 1998 and has a different married name now – said their courtship had been brief and that she had never visited Maine until after the wedding. She said he talked a lot about how important his work was and all the security clearances he had and secrets he had to keep.


In 2004, Dechaine’s defense team – which included Michaela Murphy, who is now a justice at Cumberland County Superior Court – tasked private investigators William E. Moore and Thomas Cumler with looking into Evonitz. Between them, they traveled to Spotsylvania County and Quantico, Virginia, to interview investigators and FBI profilers. They were unable to obtain solid evidence that could be presented at a trial but considered Evonitz a possible suspect and still do.

“I think it would be irresponsible not to follow up Evonitz,” Cumler said. “This is a pretty nasty crime. I’ve done probably over 100 homicides over 35 years and saw nothing like that.”

Dechaine had no prior criminal record.

“Whoever did this, this wasn’t their first time, and Dennis didn’t even have a parking ticket,” he added. “The torture – the perpetrator is getting something out of – it’s a really sick mind. Evonitz had a mind like that.”

Moore agreed. “The similarities between the Evonitz and Cherry cases are so striking,” he said. “If this new DNA technology can provide a full profile, it could help with all the suspects, either by eliminating them or confirming their involvement.”

Evonitz’s DNA profile is readily available. Enright has it in her possession and said she had shared it with Dechaine’s attorney, Nale. If new testing yields a profile, comparing it with Evonitz’s will be easy, she said: “You or I could do it.” Other alternate suspects’ DNA has also been collected and some of the profiles are in the possession of Dechaine’s supporters.



The Press Herald attempted to speak with Dechaine last week, but the Maine Department of Corrections would not allow it, citing the effect it would have on the Cherry family.

Cherry’s parents, Christopher and Debra Crossman, did not respond to interview requests.

The state Attorney General’s Office declined to participate in the story or to address the apparently exculpatory time of death evidence. The state has long argued it has the right man behind bars, a position it held during Dechaine’s unsuccessful bid for a new trial in 2015, when Gov. Janet Mills was attorney general.

The evidence used to convict Dechaine is, at first glance, extremely compelling. His alibi on the afternoon of Cherry’s abduction is that he was wandering the Bowdoin woods shooting amphetamines he had purchased in a Boston museum bathroom and lost his way back to his truck. His whereabouts can’t be accounted for until 8:45 p.m. the day of her disappearance, when he approached a Bowdoin couple’s house seeking help finding his truck. At 9:15 p.m. they left him with the police, not yet realizing they were in the area investigating a girl’s abduction.

A police photo of Dennis Dechaine’s pickup truck taken July 6, 1988, the day Sarah Cherry disappeared. Cherry’s body was discovered nearby.

His truck was found parked in a pull-out across a road and a few hundred feet away from where Cherry’s body was discovered a day and a half later. The nylon rope binding her hands appeared to have been cut from a coil in the bed of the truck. A car repair receipt and a notebook bearing Dechaine’s name were found in the driveway of the home where Cherry had been abducted. Under police interrogation, Dechaine began to question if he could have done it, though he offered no details of the crime.


But uncertainty remained. Dechaine had been a nonviolent man who could not even kill his own chickens. He had a clean criminal record and raised no red flags when psychologically examined. No fingerprints, hairs, fibers or other evidence were found on the victim or at the abduction site, and none of Cherry’s hairs, fibers or prints were found on Dechaine or in the truck he supposedly abducted her with. A police tracking dog couldn’t pick up Cherry’s scent in or around the truck. Pink synthetic fibers and other debris were found around the crime scene that couldn’t be connected to either of them, and the blood under the girl’s fingernails contained male DNA that was not Dechaine’s. Dechaine cooperated with police and, from the outset, fought for DNA testing of crime scene evidence.

His defenders argue he was framed by the real killer, who they say took items from his parked truck to implicate him. The time of death science makes this a real possibility.

During a summer heat wave, Cherry’s body was found under leaves and twigs with no sign of decomposition and with rigor mortis – the temporary tensing of muscles after death – still occurring. The original medical examiner, Ronald Roy, testified at trial that Cherry had likely been dead for 30 to 36 hours at the time he examined her, but that “it could well be longer.” His 30- to 36-hour window would have placed the time of death between 2 a.m. and 8 a.m. on July 7, at which time Dechaine had been with the police for five to 11 hours. On cross-examination, Roy also noted there was less fly activity than he would have expected on a body left in the woods in August for two days.

In 2009 Dechaine’s defense team had the medical examiner’s files and findings reviewed by two of the top forensic pathologists in the country, Walter Hofman, then coroner of Montgomery County, Pennsylvania, in suburban Philadelphia, and Cyril Wecht of the University of Pittsburgh, known for his work on the JonBenét Ramsey and O.J. Simpson cases and his criticism of the Warren Commission’s conclusions on John F. Kennedy’s assassination. They concluded the time of death had to have been later than Roy had estimated.

Hofman concluded the earliest time of Cherry’s death at noon on July 7, 16 hours after Dechaine was in custody. Wecht placed it at no earlier than 3:40 a.m. on July 7, with the most probable period between 9:40 a.m. and 3:40 p.m. that day. The relevant data, Wecht said, “does not permit an estimate of greater than (3:40 a.m. July 7) at the very most” for the time of death, underlining the last three words.

Wecht reviewed his findings for the Press Herald last week and stands by them.


“I’m not saying this to be obstinate,” he said. “It’s just a matter of science.” He also noted that all the factors that might alter rigor mortis – the heat and victim’s youth and terrorized state – would all have caused it to fade faster, not slower, pushing the time of death later.

Nale, Dechaine’s attorney, said more information on what objects will be tested will be released shortly.

Enright, whose law students continue to research cold cases Evonitz may have been involved in, said she hopes new DNA technologies will help find the perpetrators of heinous crimes and clear people who may have been falsely convicted for them.

“The idea that prosecutors and investigators hunker down on their original dumb theory is not news at all,” she said. “Sometimes it just becomes more and more impossible for them to admit they made a mistake.”

Related Headlines

Only subscribers are eligible to post comments. Please subscribe or login first for digital access. Here’s why.

Use the form below to reset your password. When you've submitted your account email, we will send an email with a reset code.

filed under: