The results of the Razumkov Centre’s latest survey show that three months before the official launch of the presidential campaign, the people have defined their preferences with respect to prospective candidates.

It is obvious that one round will not be enough, and the elections will not end on October 31st, 2004 - none of the candidates have an earthly chance to get over 50% of the vote. For several months in a row, the number of respondents unwilling to support even the highest-rated politicians has exceeded the number of their supporters. And there are no grounds to expect this correlation to change dramatically as the campaign advances.

The second round on November 14^th seems inevitable. Two candidates - “Our Ukraine” leader Viktor Yushchenko (supported by 22.6% of respondents) and Prime Minister Viktor Yanukovych (15.3%) - may get into the second round. Three months ago, Viktor Yanbukovych squeezed the Communist leader from second place, and Petro Symonenko does not seem likely to restore his lost position in the presidential candidates’ rating.

Yushchenko’s popularity is the highest, remaining practically unchanged since January 1st 2004. Yanukovych is gaining support; his rating has surpassed that of any previous head of government. At the time of Yanukovych’s appointment as Prime Minister, his presidential rating was 3%; his predecessor Anatoliy Kinakh at his prime could not count on more than 5%-6% of votes.

The diagram shows that Yanukovych’s chances to get to the second round grow markedly if he is really nominated as the single candidate from the “party of power”. In any case, the “Our Ukraine” leader has a 9% margin of safety, so far.

If Yushchenko and Yanukovych end up in the second round, they will muster 31.7% and 23% of votes, respectively. The battle will focus on the 30% of undecided, yet persuadable voters.

The national opinion poll was carried out by the Razumkov Centre Sociological Service on April 2-8, 2004. 2,020 respondents over age 18 were surveyed in 123 residential areas, including villages, urban settlements, towns and cities. Sampling error did not exceed 2.3%.