In many Zygoptera spp. females occur in different colour morphs, with one morph coloured like the male (andromorph), while the others are not (gynomorph). Two hypotheses have been proposed to explain frequency-dependent harassment of female morphs. According to the first, males should prefer the more frequent of the 2 female morphs (learned-mate recognition hypothesis). According to the second, males should prefer andromorphs more if their frequency relative to males increases, but not so for gynomorphs which always should be attempted to mate with on encounter (mimicry hypothesis). Here, it is reported on a re-analysis of earlier published data on morph-specific harassment for I. elegans, which allows examination of the 2 proposed hypotheses. The data were collected in 8 insectaries with different ratios of males and female morphs. As reported earlier, male harassment is highest on the most common female morph supporting the learned-mate recognition hypothesis. The ratio of andromorphs to males had no morph-specific effects in amounts of male harassment, wherefore the data suggest rejection of the mimicry hypothesis.