We examined efficacy of vote in multi-seats elections. Efficacy of vot
e has been calculated for single seat elections and several ways of ca
lculation have been proposed. However, research for the multi-seats el
ections is far behind. We made general considerations for the voters i
n multi-seats elections. Multi-seats elections request new ideas for t
he analysis. Especially, we concluded that the assumption that voters
have utilities on the combination of winners than the independent winn
ers is suitable. Under this assumption we calculated the efficacy for
the two-seats three-candidates case with the simplest method of multin
omial distribution model. We find that a vote is more effective in two
-seats three-candidates case than in one-seat two-candidates case. We
also find that among two-seats three-candidates elections limited vote
system (one vote for one voter) is as effective as bloc vote system (
two votes for one voter).