The aim of the present study was to develop a Type 2 diabetes symptom checklist for use in clinical and epidemiological research, which can measure differences in symptom severity between patients and detect changes over time within patients. Face and content validity of items and dimension structure were based on literature and experiences of diabetologists. Two Likert scales were used to measure symptom frequency and perceived burden. Reliability, responsiveness and validity were studied in 185 Type 2 diabetic patients. Factor-analysis confirmed the predesigned dimension structure with 34 items, distributed over 6 dimensions. The internal consistency with Cronbach alpha coefficients between 0.76 and 0.95 and test-retest reliability with Pearson product-moment correlation coefficients between 0.79 and 0.94 were satisfactory. The ability to detect change over time (responsiveness) was estimated in stable subjects, using sample size calculations. A minimal detectable mean change of 0.07-0.58 points in total score on the 10-point scale within a group of 100 subjects suggests an impressive responsiveness. Significant differences in symptom severity score were found between patients with different co-morbidity status and treatment modes, indicating satisfactory construct validity of the dimension structure. The Type 2 Diabetes Symptom Checklist was found to be a useful diabetes-specific symptom severity assessment method for clinical and epidemiological studies.