This paper reports on two studies, each concerned with sex dierences in the estimates of Gardner's seven basic types of intelligence'. In the ยฎrst study, 180 British adults were asked to estimate their own intelligence on the seven intelligence factors. Only one (mathematical/logical) showed a signiยฎcant sex dierence, with males believing they had higher scores than females. Factor analysis of these seven scales yielded three interpretable higher-order factors. There was a similar sex dierence on only one factor (mathematical/spatial intelligence), which showed males rating themselves higher than females. In the second study, 80 student participants completed the same seven estimates of intelligence, plus a standard sex-role inventory, in order to separate sex and sex role in the self-estimation of intelligence. A series of sex ร sex-role ANOVAs showed some eects, particularly for mathematical, musical, and spatial intelligence, but nearly always for sex and not sex role. Results suggest that previous studies which found consistent sex dierences in self-estimates of overall intelligence (g') may have over-exaggerated the issue as the dierence is clearly conยฎned to a limited number of factors of intelligence.