Objective: To analyze how the cognitive, affective, and somatic-vegetative symptoms of depression included in the modified Zung Self-Rating Depression Scale (ZSDS) relate to pain severity. Subjects: Three groups of subjects were compared. The “no pain” group (n = 64) reported no musculoskeletal symptoms during the preceding year, the “low pain” group (n = 119) had a pain rating of ≤4, and the “high pain” group (n = 137) had a rating of ≥5 on a numerical pain rating scale (0–10). Methods: Sum scores of the ZSDS and means of separate items in the three subgroups were compared. Factor analyses were performed to evaluate the structure of the ZSDS, and differences between subgroups on weighted factor scores were analyzed. Results: The “high pain” group differed statistically significantly from both other groups on the sum score and on most of the items. Comparison of weighted factor scores in a three-factor solution revealed that in Factor I, “Psychic Energy,” the “high pain” group differed from both other groups. In Factor II, “Despair,” no differences between groups were found, while in Factor III, “Sleep problems,” statistically significant differences were found between all subgroups; more severe pain was associated with more sleep problems, fatigue, and restlessness. Conclusion: ZSDS items describing symptoms which could be a consequence of the pain problem (e.g., sleep problems) are clearly related more to pain severity than other indicators of depression—e.g., feelings of hopelessness. A diagnosis of depression based on a sum score of an inventory which contains somatic-vegetative signs of depression may lead to an overestimation of depression in chronic musculoskeletal pain patients.