To evaluate the reliability and concurrent and discriminant validity of the Problem Areas in Diabetes (PAID) scale, a new measure of emotional functioning in diabetes. A battery of questionnaires, including the PAID, was completed by 256 volunteer diabetic outpatients. In our analyses, we examined the PAID's internal structure and compared mean IDDM and NIDDM treatment group scores in regression analyses to explore its discriminant validity. We also evaluated concurrent validity from the correlations between the PAID and diabetes-specific measures of coping and health attitudes and HbA1c. Principal component analyses identified a large emotional adjustment factor, supporting the use of the total score. Significant sizable correlations were found between the PAID and a range of selected health attitudinal measures. There were significant differences (with small-to-moderate effect sizes) in PAID scores between IDDM and NIDDM patients and between IDDM and NIDDM insulin- and tablet-treated subgroups; no differences were found between NIDDM insulin- and tablet-treated subgroups. The study findings provided support for the construct validity of the PAID, including evidence for discriminant validity from its ability to detect differences between IDDM and NIDDM treatment groups expected to differ in the emotional impact of life with diabetes. Future studies should explore the PAID's performance in nonspecialist treatment settings as well as its responsiveness to clinical change.