This study was designed to develop an instrument for measuring attitudes toward pharmacist-physician collaborative relationships for administration to practicing pharmacists and physicians, as well as to students in pharmacy and medical schools. Based on a review of literature, a preliminary version of an instrument was developed (30 items), and through a pilot study of face validity and content validity with 12 pharmacists and 10 physicians, 18 items were chosen for quantitative analyses. We asked 88 respondents (61 pharmacists, 27 physicians) to judge the relevance, clarity, and representativeness of each item to the concept of pharmacist-physician collaborative relationships. Sixteen items with a relevancy endorsement greater than 85% and significant item-total score correlations were retained. The following underlying constructs emerged from factor analysis: "collaboration and team work," "accountability," "overlapping responsibility," and "authority". These factors supported the multidimensionality and construct validity of the instrument. No gender difference was observed; however, pharmacists scored higher than physicians on the total score of the instrument. The Cronbach's coefficient alpha was .81 for pharmacists, .92 for physicians, and .87 for the combined sample. Encouraged by these preliminary findings, we plan to undertake further research to examine the instrument's psychometric properties including criterion-related and predictive validities with larger and more representative samples of pharmacists, physicians, and students in pharmacy and medical schools.