We investigated male sexual behaviour and the cost of sexual harassment, as measured by the reduction of female feeding time in the presence of a male, in a cave-dwelling population of Poecilia mexicana, in which sexual harassment does not occur naturally. We asked whether the lack of sexual harassment in this population is due to low sexual activity of the males, or low feeding motivation of the females. We experimentally increased the sexual activity of males or the femalesrsquo feeding motivation, or we used a combination of both treatments. Female feeding time was not lower in the presence of a male than in the presence of a female after sexual deprivation of the males or food deprivation of the females. Only in the combined experiment was female feeding time lower in the presence of a large male than in the presence of a small male, indicating a weak effect of sexual harassment by large males. Virgin females did not suffer a cost of sexual harassment, indicating that sexual experience does not cause the lack of sexual harassment in cave mollies. Males from a surface population, where sexual harassment occurs, significantly reduced the feeding time of cave-dwelling females even though these males exhibited surprisingly little sexual behaviour. The sexual activity of cave mollies did not correlate with male body size in any experiment, indicating that even after sexual deprivation, small cave molly males do not switch to the ldquoalternativerdquo mating behaviour known in surface-dwelling P. mexicana, where sexual activity is correlated negatively with male body size.