The study was undertaken to investigate the test-retest stability and the criterion-related validity of a modified Swedish version of an exercise motivation index (EMI), and its use with individuals with rheumatic conditions, and with healthy individuals who exercised regularly. The EMI consists of 23 statements divided into three sub-scores for physical, psychological and social motivation. Ninety-five individuals with rheumatic conditions (mean age 60 years, mean symptom duration 15 years, 79% female) and 131 healthy individuals (mean age 52 years, 76% female), all attending exercise classes at least once a week, filled out the EMI. Sub-samples also filled out three visual analogue scales designed to measure physical, psychological and social exercise motivation, and filled out the EMI a second time one week later. The results indicated that psychological and physical exercise motivation was equally important in both samples. Social motivation was less important in both samples, but more pronounced in the rheumatic sample and among older individuals. In the rheumatic sample, physical motivation was more important among women and psychological motivation was more important among younger individuals. Test-retest stability for the EMI was satisfactory in both samples, while criterion-related validity was poor. The results of our preliminary investigation of the EMI suggest that the survey of physical, psychological and social motivation for exercise seems meaningful. Further work on the validity of the EMI is needed.