Aims: To assess the relationship between pain and HbA1c levels in a predominantly black population with diabetes, and to determine whether self-management behaviours (exercise and diet) and symptoms of depression mediate this relationship. Methods: We analysed cross-sectional data from 417 community-dwelling individuals with diabetes in rural Alabama, USA. Binary logistic regression was used to analyse the relationship between pain and HbA1c levels, defined as relatively good [≤ 64 mmol/mol (≤ 8.0%)] and relatively poor [> 64 mmol/mol (> 8.0%)], after adjusting for sociodemographics, insulin use, medication count, cigarette smoking history and body mass index (BMI). We examined the mediating roles of exercise, diet, and symptoms of depression using bootstrapping. Results: Participants were primarily black (86.6%), female (76.1%) and reported an annual income of ≤$20,000 (52.7%). Their mean (sd) age was 59.6 (12.8) years. The majority of the participants reported moderate to extreme pain (71.5%). Participants reporting pain were more than twice as likely to have HbA1c levels > 64 mmol/mol (8.0%) in the fully adjusted model (odds ratio 2.33 [95% CI 1.28-4.24]; P < 0.05). Diet significantly mediated the relationship between pain and HbA1c control (β = 0.06; 95% CI: 0.01-0.17), but only in the unadjusted model. Exercise and symptoms of depression were not significant mediators. Conclusions: A significant independent relationship between pain and HbA1c control was found in this mainly black population, which was not explained by self-management behaviours or symptoms of depression. Future research is needed to delineate the mechanism by which pain influences HbA1c control, especially among black people with diabetes on low incomes. © 2013 The Authors. Diabetic Medicine © 2013 Diabetes UK.