To clarify the effects of noise, sex, and intelligence on student performance, 289 sixth-grade students were randomly assigned either the Standard Progressive Matrices (SPM) or the STEP Reading Test Form 3 (STEP III) to be taken under high- or low-noise classroom conditions, with gender and intelligence as variables. Students who took the SPM while listening to low levels of previously recorded classroom noise performed better than did students in the high-noise-level group. When below-average intelligence students took the STEP III under high-noise conditions, their test scores improved, while above-average intelligence students performed best on STEP III when noise levels were low. Students with above-average intelligence did significantly better on both tests, in both noise settings. This study supports the contention that the effect classroom noise has on student performance depends on the task involved, but does not support previous findings that suggested sex was a significant variable in determining student performance under various noise-level conditions. (Author/DCS)