The development and validation of a four-item scale, the ATAPS, measuring attitude toward working with arthritis patients is described. Three separate samples of occupational and physical therapists were surveyed. Data from the first random sample were used to select a reliable subset of items from a larger pool and to provide a baseline for comparison with the second sample. The second sample comprised arthritis workshop attendees and confirmed the scale's reliability and its ability to differentiate between therapists with an active interest in arthritis and a group with no special interest in the disorder. The third stratified random sample further confirmed the scale's reliability and demonstrated that the ATAPS scores were associated with the number of arthritis patients in therapists' practices but not with measures of professional experience unrelated to arthritis.