The level of homosexual behaviour is evaluated in one laboratory population of seed beetle and derived lines selected to reproduce early (E) or late in life (L), where inadvertent selection for either low or high heterosexual activity has been detected. The magnitudes of homosexual interaction, measured as chasing and mounting individuals of the same sex, are estimated over different age classes. These magnitudes are correlated with previously observed levels and patterns of age-specific variation of heterosexual activity of both sexes in the E and L experimental lines. The results obtained support the perception error hypothesis proposing that a low degree of sexual discrimination is genetically correlated with high sexual activity. The fitness costs of the same-sex interactions are tested by assessing their effects on longevity. In both sexes, the longevities of homosexual pairs are reduced relative to individually-housed virgin beetles in both the E and L lines, although homosexual interactions have a more pronounced effect on male survival than on female survival. Although the results obtained suggest that the longevity cost of homosexual interactions can be substantial, this cost is much smaller than the cost of heterosexual interactions.