The poll found that among registered voters, 25 percent said they had heard "a lot" about Rev. Jeremiah Wright's comments, while 33 percent said they had heard some. Forty-two percent said they hadn't heard about the comments.
Of those voters, sixty-five percent said it didn't make a difference in their view of Obama. However, of those whose opinion is changed, the net impact is very negative. Thirty percent said it made them have a less favorable view, whereas 2 percent said it made their view more favorable.
Interviewing for these questions was completed Sunday and Monday, before.
Democrats are especially apt to say their views are unchanged, with 76 percent saying it has made no difference of their view of Obama, 15 percent saying it made their view less favorable and 2 percent saying it made their view more favorable.
See The Complete Poll Results
Video: Obama On Anger Between Races
Republicans are the most likely to say their view has been affected: 47 percent say they've become less favorable, and 53 percent said it did not make a different.
Sixty-one percent of independent voters say they are unaffected, but 36 percent said it made their view less favorable. Two percent of independents said it made them more favorable view.
Overall, unfavorable views of Obama are up somewhat from February. His favorable ratings remain largely unchanged at 44 percent, but there has been some movement from undecided views to unfavorable views, from 23 percent in February to 30 percent now.
Among Democratic primary voters, unfavorable views have doubled, but favorable views remain dominant: more than six in ten Democratic primary voters are positive.
Independent voters - a group Obama has successfully courted in many primaries and would try to draw in the general election - still view him favorably, about the same as last month. Unfavorable views have risen among Republicans.
This poll was conducted among a random sample of 844 adults nationwide, including 749 registered voters, interviewed by telephone March 15 - 17, 2008. Interviewing for the questions on Wright and Obama was conducted March 16-17, 2008 among a sample of 355 registered voters. The error due to sampling for results based on the entire sample could be plus or minus three percentage points and five points for those questions. The error for subgroups is higher.