In a problem-based curriculum, emphasis is placed on the groups in which students learn to analyze problems and to contribute to the solution of a problem. This paper describes an instrument that aims to measure individual group performing and presents some psychometric results. Reliability and validity were studied with 240 students in groups of 9 with 1 tutor. The instrument appears to be reliable, with coefficient alpha varying from 0.84 to 0.92, and it seems that it is a valid measure for characteristics that are important in a problem-based curriculum: (1) participating in discussions; (2) explaining things to other students; and (3) a broad study approach. (Contains 2 tables and 11 references.) (SLD)