It is unclear whether depressive symptoms are a risk factor for incident diabetes or diabetes is a risk factor for depressive conditions. Therefore, we examined the longitudinal bidirectional associations between depressive symptoms and type 2 diabetes (T2D) as well as the impact of both diseases on (all cause) mortality in a sample of primary care patients over a 3.5-years follow-up period on average. Depressive symptomatology, defined by the Depression Screening Questionnaire (DSQ), was examined both categorically and dimensionally. Patients were categorized as normal fasting glucose (NFG), impaired fasting glucose (IFG), and T2D (untreated, oral antidiabetics, insulin/combined treatment) according to physician ratings and baseline lab values. Incidence rates of T2D were 25.6 and 20.9 per 1000 person-years for those with and without depressive symptoms, respectively. The unadjusted risk of incident type 2 diabetes was 1.03 times higher (CI(95%): 1.01-1.06) for each 1-point increment in DSQ score. The incidence rates of elevated depressive symptoms per 1000 person-years were 30.5 for NFG, 34.2 for IFG, 36.4 for untreated T2D, 32.3 for oral treated T2D, and 47.8 for insulin/combined-treated T2D patients. Compared to NFG patients, insulin-treated patients had a higher risk of incident depressive symptoms (HR: 1.71; CI(95%): 1.03-2.83) and oral-treated patients had a lower risk (HR: 0.58; CI(95%): 0.36-0.96). Higher mortality rates were associated with both diseases compared to patients without T2D or depressive symptoms at baseline (HR: 2.49; CI(95%):1.45-4.28). Results indicate that especially insulin treatment in T2D patients is associated with incident depressive symptoms.