A day after the Justice Department released hundreds of documents purporting to link Bruce E. Ivins to the 2001 anthrax killings, scientists and legal experts criticized the strength of the case and cast doubt on whether it could have succeeded.

Federal investigators presented a raft of circumstantial evidence this week intended to prove Ivins' guilt beyond a reasonable doubt. But officials lacked direct evidence, such as hair fibers, DNA samples or handwriting analysis, that the eccentric microbiologist created the deadly powder in his Fort Detrick lab. Questions also remain about Ivins' ability to convert the spores stored in his lab into the powder sent through the mail.

More than half a dozen experts in law and bioterrorism pointed out yesterday what they consider major flaws in the government's case and said they were not convinced that Ivins acted alone in mailing the letters that killed five people - or that he was involved at all. They said that the science that led the FBI to Ivins has not been explained and that the other evidence did not amount to conclusive proof.

Because Ivins committed suicide last week, that evidence will never be tested at trial, but his attorney has repeatedly insisted that the scientist was innocent.

The FBI said it used a sophisticated mapping technique to connect the anthrax in the letters with a flask in a Fort Detrick lab where Ivins worked. But that technique is so new that in the hands of a skilled defense lawyer, it could be "unraveled in front of a jury," said Michael Greenberger, a professor at the University of Maryland Law School.

Not all legal experts were skeptical of the case. Former federal prosecutor E. Lawrence Barcella said the FBI appeared to have done a remarkably thorough investigation. "They've made a very strong circumstantial case, an extremely strong circumstantial case," Barcella said.

Others said the focus on Ivins' lab raised concerns. The government said that 16 government, commercial and university labs had the strain of anthrax with the same genetic mutations as the anthrax used in the attacks. Only one of those 16 was in Maryland or Virginia - where the government thinks the envelopes used in the attacks were purchased. That lab is the one where Ivins worked.

"I thought that was a bit of a stretch," said Jonathan D. Tucker, a biological warfare expert on a federal commission to prevent terrorism and the spread of weapons of mass destruction. "It's such a key piece of their argument, and it's based on inference. We don't know which other labs had the strains with the mutations."

And even at Fort Detrick, the government said that more than 100 people had access to the flask, creating a lot of room for reasonable doubt.

"There is a classic defense mechanism - to raise reasonable doubt by presenting the jury with viable options as to how the crime was perpetrated," Greenberger said. "This case would have been a very, very difficult case to prove."

The FBI defended its work Wednesday, saying the painstaking investigation could lead to only one conclusion: Ivins was solely responsible for the attacks. They said the fact they had only circumstantial evidence did not weaken their case, which they now consider closed.

"Thousands of prosecutors in thousands of courthouses across this country every day prove cases beyond a reasonable doubt using circumstantial evidence," Joseph Persichini, assistant director of the FBI's Washington field office, said Wednesday.

In addition to the flask connection, Barcella said the time logs that show Ivins used his lab late at night just before the anthrax mailings raise strong suspicions. He said prosecutions based solely on circumstantial evidence were common and that the strongest such cases relied on the premise that all other alternative scenarios have been ruled out.

That's what happened with the anthrax case, Barcella said. The FBI "eliminated every possibility until they were left with one, which is a tedious but very thorough way to do an investigation," he said.

But other experts say the FBI has not shown how it cleared the 100 or more other people who had access to the flask - the kind of detail that would be forced out at trial. Lacking a trial in this case, experts said Congress should hold hearings or order an independent review of the evidence.

"If those questions are not resolved, there will always be a residue of doubt that the perpetrator will still be at large and that an innocent man may have been accused," said Tucker, of the federal anti-terrorism commission. "It's very important to tie up these lingering loose ends and address the gaps."

Among the unresolved questions:

•How do officials believe Ivins made the anthrax? The FBI says Ivins used his lab to convert anthrax spores into powdered anthrax, but no proof has been presented that he had the equipment or the expertise to do so.

"I'm waiting for it to be shown that the quantity and the quality of the powders in the anthrax letters could have been produced in those suites" at Fort Detrick, said W. Russell Byrne, who retired from Fort Detrick in 2003 and was Ivins' supervisor from 1998 to 2000. "I don't know how to make the stuff," he said.