This study investigated whether adapting instruction to gender-type preferences can improve the attitudes of students, especially female students, and their performance in mathematics. Subjects were 69 young adult college students in a remedial mathematics program whose interest in gender-type activities was determined. Students were randomly assigned to gender-adapted or gender-maladapted treatment conditions. In the gender-adapted condition, females received the feminine version of instruction, and males received the masculine version; in the maladapted condition, students received instruction designed for the other gender. The absence of performance differences between conditions supports previous research that found no effect on performance of problem context sex stereotype. This result, coupled with improved mathematics ability scores across conditions, suggests that dynamic worked-example mathematics instruction improves students' performance regardless of the story context in which it is set. Differential results in participants' attitudes, however, reveal the potential impact adapted instruction may have on internalized beliefs and attitudes about mathematics. (Contains 13 Power Point slides and 10 references.) (SLD)