Liberal Justin Trudeau may face a difficult trip with Manitoba voters if the number of new polls persists during the next federal election, now less than a year away.
The Probe Research poll found support for federal Liberals had fallen 11 points in Manitoba, and was 14 percent full in cleaning outside Winnipeg, since the 2015 federal election.
The survey reached 1,105 Manitobans randomly between November 27 and December 6.
Forty-two percent of Manitoba plans to choose the Tories, according to the poll, while 34 percent say they will vote for Liberals. Sixteen percent of voters said they would support the NDP, and support for the Canadian Green and People's Party each received five and two percent.
While the poll found that Liberals continued to lead the Conservatives in Winnipeg, support for the ruling party had dropped from 53 percent in 2015 to 43 percent. The poll found 33 percent of Winnipeggers' plans to choose the Tories.
Support on the Perimeter Highway is the strongest in the core, where 46 percent of respondents said they would vote for Liberals, 24 percent planned to vote for the NDP and 18 percent said they planned to vote for the Conservatives.
The Liberals also maintained significant leadership over the Conservatives in the northwest and southwest of the city, but both parties were statistically bound in the southeast and northeast.
Outside Winnipeg, voters told the poll that they would not vote for Liberals.
The survey found only 19 percent of voters planning to choose Liberals in the countryside, down from 33 percent in 2015, while 58 percent of voters decided to say to voters that they would support the Conservatives this time, up from 48 percent in the last election.
The survey also found 48 percent of male voters in Manitoba planned to vote for the Conservatives, compared with 31 percent who said they would vote for Liberals.
On the other hand, women voters appear to be divided between the two parties, with 37 percent promising Liberal support, and 36 percent planning to elect Conservatives.
The investigation was carried out on the field between November 27 and December 6 this year. This random sample size will produce a margin of error +/- 2.9%, 19 times that of 20. The error margin is higher in each survey sub-group.