Poll: Giuliani Leads, But Support Shaky

Former New York City Mayor Rudy Giuliani maintains his status as the most-preferred nominee among Republican primary voters nationwide, a new CBS News poll finds. But six in ten of his backers say they have reservations about the candidate.

Twenty-nine percent of Republican primary voters favor Giuliani, while 21 percent support former Senator Fred Thompson. Senator John McCain's support remains at 18 percent, the same as last month, while former Mass. Gov. Mitt Romney receives 12 percent support. Former Arkansas Gov. Mike Huckabee, included in the choices in the poll for the first time, garners 4 percent support.

Giuliani leads among moderate primary voters, while Thompson leads among white evangelicals. Among conservatives, Giuliani and Thompson are neck and neck at 24 percent each. McCain runs second among moderates behind Giuliani, and third among conservatives and white evangelicals.

Giuliani's backers have growing reservations about their candidate, however. Today, 60 percent say they have their doubts, up from 45 percent in August. Thirty percent strongly favor the former New York City mayor, down from 47 percent two months ago.

Twenty-seven percent of Republican primary voters would definitely vote for Giuliani in the general election if he became the Republican Party's presidential nominee, while 11 percent say they definitely will not vote for him. Fifty-one percent of Democrats, meanwhile, definitely would not vote for Giuliani, while 18 percent would at least consider it. A majority of voters overall say they would either consider voting for him or that it is too early to say.

All the major candidates for the Republican nomination are seen positively by Republican primary voters, although Giuliani is the only one with a net favorable rating from registered voters overall. Many voters have yet to make up their minds, particularly when it comes to Thompson, Romney, and Huckabee.

Click here to view the complete poll results.

This poll was conducted among a random sample of 1282 adults nationwide, interviewed by telephone October 12-16, 2007. The error due to sampling for results based on the entire sample could be plus or minus three percentage points. The error for subgroups is higher.