People can estimate their own and their romantic partner's intelligence (IQ) with some level of accuracy, which may facilitate the observation of assortative mating for IQ. However, the degree to which people may overestimate their own (IQ), as well as overestimate their romantic partner's IQ, is less well established. In the current study, we investigated four outstanding issues in this area. First, in a sample of 218 couples, we examined the degree to which people overestimate their own and their partner's IQ, on the basis of comparisons between self-estimated intelligence (SEI) and objectively measured IQ (Advanced Progressive Matrices). Secondly, we evaluated whether assortative mating for intelligence was driven principally by women (the malescompete/females choose model of sexual selection) or both women and men (the mutual mate model of sexual selection). Thirdly, we tested the hypothesis that assortative mating for intelligence may occur for both SEI and objective IQ. Finally, the possibility that degree of intellectual compatibility may relate positively to relationship satisfaction was examined. We found that people overestimated their own IQ (women and men 30 IQ points) and their partner's IQ (women = 38 IQ points; men approximate to 36 IQ points). Furthermore, both women and men predicted their partner's IQ with some degree of accuracy (women: r = 0.30; men: r = 0.19). However, the numerical difference in the correlations was not found to be significant statistically. Finally, the degree of intellectual compatibility (objectively and subjectively assessed) failed to correlate significantly with relationship satisfaction for both sexes. It would appear that women and men participate in the process of mate selection, with respect to evaluating IQ, consistent with the mutual mate model of sexual selection. However, the personal benefits of intellectual compatibility seem less obvious.