The online survey of likely voters found Obama and Romney were each supported by 46 percent of the electorate as they prepare for their final televised debate on Monday night (local time).
Romney trailed by 1 percentage point when the poll was last published on Saturday. The two candidates have remained within three percentage points of each other since shortly after their first debate on October 3.
"Today's number emphasizes the fact that the race is very close. We enter the final debate with the candidates literally neck and neck," Ipsos pollster Julia Clark said.
However, Obama still holds a substantial advantage in the battleground states that will determine the outcome of the election. Ipsos projects Obama will carry hotly contested states such as Florida, Ohio and Virginia, for a relatively comfortable electoral college victory.
Reuters/Ipsos online polls are measured using a credibility interval. The survey of 957 likely voters, conducted between October 18 and October 22, has a credibility interval of 3.6 percentage points.
Among the larger pool of registered voters, Obama leads Romney 45 percent to 42 percent.
Early voting is now under way in 49 states and the Distinct of Columbia, and the poll found that 16 percent have already cast their ballots.
While the election may be already over for millions of voters, for others it remains very much an open question. Among likely voters, 6 percent said they did not know which candidate they would vote for, or refused to say.
Among those who have settled on a candidate, 15 percent of registered voters said they could still change their minds.
Monday's debate in Boca Raton, Florida, represents the last chance either candidate will likely get to make their case directly to a television audience of millions of voters.
The debate is unlikely to change the dynamics of the race substantially unless either Obama or Romney makes a major mistake, Clark said.
"I think this is how we're going to see the numbers looking for the next two weeks. It's just very close," she said.
