Patient satisfaction is associated with positive diabetes outcomes. However, there are no identified studies that evaluate both patient- and clinic-level predictors influencing diabetes care satisfaction longitudinally. Data from the INtegrating DEPrEssioN and Diabetes treatmENT trial was used to perform the analysis. We used fixed and random effects models to assess whether and how changes in patient-level predictors (treatment assignment, depression symptom severity, systolic blood pressure, body mass index, LDL cholesterol, and haemoglobin A1C) from 0 to 24 months and clinic-level predictors (visit frequency, visit cost, number of specialists, wait time, time spent with healthcare provider, and receiving verbal reminders) measured at 24 months influence diabetes care satisfaction from 0 to 24 months. Model 1 (patient-level predictors) accounted for 7% of the change in diabetes satisfaction and there was a significant negative relationship between change in depressive symptoms and care satisfaction (β = -0.23, SE = 0.12, p < 0.05). Within Model 1, 2% of the variance was explained by clinic-level predictors. Model 2 included both patient- and clinic-level predictors and accounted for 18% of the change in diabetes care satisfaction. Within Model 2, 9% of the variance was attributed to clinic-level predictors. There was also a cross-level interaction where the change in depression had less of an impact on the change in satisfaction for those who received a verbal reminder (β = -0.11, SE = 0.21, p = 0.34) compared with those who did not receive a reminder (β = -0.62, SE = 0.08, p < 0.01). Increased burden of depressive symptoms influences diabetes care satisfaction. Clinic-level predictors also significantly influence diabetes care satisfaction and can reduce dissatisfaction in primary care, specifically, reminder calls from clinic staff.