The performance of 18 boys and 18 girls on four problem-solving tasks set in science contexts was compared. The tasks were administered in a one-to-one testing situation and assessments were made by direct observation, questioning, and by using written records. The tasks were valid and reliable, and the samples of boys and girls were matched for ability and curriculum background. Past studies have identified gender differences in performance on science tasks; however, this study found little evidence to support these findings. Few significant differences in performance were found. No gender differences were detected in observation, reporting, or planning skills, and there was no differential performance on the use of scientific language. Girls performed less well in relation to self-reliance, and performance differences on the interpretation skill approached significance with boys' performance superior.