The functions of the science supervisor were studied and an instrument to evaluate science supervisor practices was developed. Supervisory functions were categorized into six areas. Case studies concerning problems in each area were written. Five solutions to each problem were developed. A jury of science educators and science supervisors was selected to respond to the case studies. The paired comparison technique was used to analyze the data. The responses of the science educators and the supervisors were compared and of the ten case studies in the instrument, seven yielded statistically significant data. The jury consistently indicated that dogmatic dictatorial methods were the least attractive of the alternatives. They believed supervisors functioned best using indirect methods. These conclusions agree with what appears in the professional literature. The instrument is recommended by the investigator for use as resource information, for evaluating supervisory services, for evaluating applicants for supervisory positions, and for a basis of research in science supervision. (BC)