A field experiment was conducted in the Matsumae area of Hokkaido, Japan, during June and July 2002, to investigate the effects of different entrance designs on the catch efficiency of fish traps by fishing with commercial traps (entrance inclination angle [ α ] = 37 ° ; funnel length of entrance [ L f ] = 22 cm) and experimental traps. The experimental traps were of the same size and similar design as commercial traps, with different entrance inclination angles (trap E1: α = 46 ° ; E2: α = 27 ° ; E3: α = 0 ° ; all L f = 22 cm) or funnel lengths (E4: α = 37 ° , L f = 8 cm). In total, 2200 fish during 200 trap hauls were captured. The catch was significantly higher using both traps E2 and the commercial trap than with trap E3 ( P < 0.05), and the catch of trap E2 was higher than that of the commercial trap. There were no significant differences in mean fish body length or the frequency distributions of body length among trap types (E1, E2, E3 and commercial). The funnel length of the entrance also affected the catch of traps. Trap E4 had significantly higher catches than the commercial trap ( P = 0.04) when traps were deployed for a 1-day soak time. Fish body length frequency distributions did not differ between trap E4 and the commercial trap. The results showed that catch can be greatly affected by trap entrance designs.