Interviewing for the 2004 exit polls was the most inaccurate of any in the past five presidential elections as procedural problems compounded by the refusal of large numbers of Republican voters to be surveyed led to inflated estimates of support for John F. Kerry, according to a report released yesterday by the research firms responsible for the flawed surveys.

The exit pollsters emphasized that the flaws did not produce a single incorrect projection of the winner in a state on election night. But "there were 26 states in which the estimates produced by the exit poll data overstated the vote for John Kerry . . . and there were four states in which the exit poll estimates overstated the vote for George W. Bush," said Joe Lenski of Edison Media Research and Warren Mitofsky of Mitofsky International.

The polling firms presented their findings in a much-anticipated report to the sponsors of the Election Day surveys, a consortium of news organizations that includes ABC, CBS, NBC, Fox, CNN and the Associated Press.

Throughout election night, the national exit poll showed the Massachusetts senator leading President Bush by 51 percent to 48 percent. But when all the votes were counted, it was Bush who won by slightly less than three percentage points. Larger discrepancies between the exit poll estimates and the actual vote were found in exit polls conducted in several states. At the request of the media sponsors, Mitofsky and Lenski are continuing to examine exit polling in Ohio and Pennsylvania, two critical battleground states where the poll results were off.

The differences between the final exit poll results and the vote count revived criticisms of the exit polls fueled by consecutive election-night debacles in 2000 and 2002. They also fueled assertions that the exit poll results were accurate and that it was the vote count that was flawed or deliberately manipulated to deliver the election to Bush.

The analysis found no evidence of fraud resulting from the rigging of voting equipment, a contention made repeatedly by those who question the 2004 vote.

Lenski and Mitofsky compared the exit-polling results with the final vote tally in 1,460 precincts where interviews were conducted and vote returns were available.

"Our investigation of the differences between the exit poll estimates and the actual vote count point to one primary reason: in a number of precincts a higher than average within-precinct error most likely due to Kerry voters participating in the exit polls at a higher rate than Bush voters. . . . While the size of the average exit poll error has varied [in past elections], it was higher in 2004 than in previous years for which we have data," Lenski and Mitofsky wrote.

But they acknowledged in the report that they remain at a loss to explain precisely why Bush supporters, or Republicans generally, were more likely to refuse to be interviewed than Kerry voters.

Their investigation identified other factors that contributed to errors in the 2004 exit polls. Interviewing in precincts where polltakers were required to stand farther away from the polls were less accurate than those where interviewers had easier access to voters leaving the polling places. Poor weather conditions also pushed down cooperation rates. They suspected that there were more young people working as interviewers in 2004, which they said was another potential source of error.

Adding to the confusion, programming errors were discovered and corrected in the afternoon of Election Day, and a technical problem severely disrupted access to the system for nearly two hours late on election night.