This exploration of the potential for gender-role stereotypes to interfere with the constructive integration of computers into elementary and secondary school classrooms suggests that gender-related differences in perceptions of abilities in mathematics and science are the precursors of similar perceptions regarding computer use. Literature pertaining to the development of gender-role perceptions and the effects of these perceptions on self-image and achievement is examined, including studies concerning attribution theory, and the influence of context on learning is discussed. Methodological constraints in gender-role research are considered, including problems with prior research resulting from lack of adequate sampling during test construction, poorly defined constructs, simplistic and biased conceptual bases, instrument bias (even in newer scales), and lack of comparability across instruments. Directions for the conceptualization of new research are also described. A 29-item reference list is included. (Author/LMM)