An experiment was conducted to examine whether word initial pitch accent information could be exploited to reduce possible word candidates by speakers of an accentless dialect in Japan. 40 native high school students from Fukushima were presented with Tokyo Japanese materials used in an earlier study, employing a gating task. Results show that the subjects performed significantly above chance, but their responses showed less sensitivity to the information in the input, and greater bias toward vocabulary distribution frequencies, than had been observed with Tokyo Japanese listeners. The whole response pattern was identical to that of speakers of Kumamoto dialect which is another accentless dialect in our earlier study. The results suggest that one of the main features of accentless dialects may be characterized by less effective use of word initial pitch information to reduce possible word candidates.