A method is described for evaluating the performance of light traps. The mosquito collections of light traps operated within a few feet of each other showed important differences, especially when the mosquitoes were present in large numbers. These differences were eliminated when the traps constantly changed position. There was no evidence that moving traps caught fewer mosquitoes than stationary traps. Lighted traps caught more mosquitoes than unlighted traps although the latter caught substantial numbers of Culex but not Aedes mosquitoes. The light intensity of the bulb illuminating a trap was directly related to the numbers of individuals of a species taken by that trap, except in the case of C. p. quinquefasciatus females. In fact the number of mosquitoes appeared to be directly proportional to the light intensity, with the exception noted. No repellency was noted at the highest intensities tested. Results with C. p. quinquefasciatus females were not yet clear-cut but females of this species appear to show a low attractance to light on dark nights but not on light ones. The effect of moonlight on females of this species is so pronounced that unlighted traps probably give a much better indication of abundance than lighted ones.