Children from two elementary schools, one with thirty-three percent male teachers and the other with no male teachers, were asked to attribute a series of descriptive statements regarding teacher competence to a hypothetical male or female teacher pictured in a cartoon drawing. Students with male teachers were found to make significantly more non-stereotyping attributions than students with no male teachers (F (1,183)=15.07,p<.001). Multivariate analysis (MANOVA) was carried out on nine competency subscales and for a general competence composite score with school and subject gender as factors. No significant differences between schools were found but significant effects of sex and of a sex by school interaction were found. A discriminant analysis uncovered two canonical variables that discriminate among the four groups: female treatment, male treatment, female control, male control. The first function, number of female selections overall, discriminated the females from the males. The second canonical discrimination function, attribution of mismanagement, revealed that the responses of boys and girls are not the same from school to school. Control school children made negative attributions in the mismanagement subscale to opposite sex teachers, while treatment children made negative attributions to same sex teachers. In the assignment of mismanagement items, the presence of male teachers significantly influenced responses which ran counter to theories of sex-role identification, including projection by elementary school age children of negative qualities to the opposite sex. An analysis of descriptive data indicates that boys with male as well as female teachers made the most egalitarian attributions to male and female teachers. Data suggest that the inclusion of men as members of the elementary school faculty could be most beneficial to boys. The presence of male teachers was not found to influence boys' interest in becoming teachers. Interest in teaching was found to differ significantly by sex (X2=24.036, df=1,183, p<.001) but not by school. Neither control school nor treatment school boys wanted to become teachers, while girls overall reported significant interest in teaching careers. Limits in generalizability of the study due to differences in the two schools, a college laboratory school and a Catholic parochial school, are discussed. Similar patterns in responses of girls from the two schools and of boys from the control school suggest that parental and curricular differences of the two schools cannot of themselves account for significant differences.