Genital morphology shows peculiar patterns of variation among insect species. Traditionally, genital species specificity has been assumed to serve as a mechanical isolation system between species (the lock-and-key hypothesis). Most recent studies suggest, however, that such variation may also be because of sexual selection. These two hypotheses give different predictions on genital variation within and between species. We tested the lock-and-key hypothesis by morphometrically exploring variation and allometry in male genitalia in three closely related Euxoa moth species. Single genital characteristics usually show overlap between species. As a whole, internal genitalia distinguish species better than external genitalia. The size of genitalia is generally correlated with body size, but the relationship is strongly negatively allometric so that the size of internal genitalia increases least with body size. These findings support the lock-and-key hypothesis. Both external and internal male genitalia show morphometric variation both within and between species, but the variation is significantly smaller in external genitalia. As internal genitalia are assumed to work as ‘keys’ in moths, this finding does not support the predictions of the lock-and-key hypothesis. Therefore, we cannot unambiguously support or reject this hypothesis. Our results agree well with the sexual selection hypotheses, particularly, the one-size-fits-all concept of the cryptic female choice hypothesis, which suggests stabilizing selection on genital size, but at the same time allows genital shape to vary relatively considerably.