Objective The objective of this study was to assess the risk of diabetes in older adults with elevated depressive symptoms, prediabetes, or both. Method This study included 4129 participants from the English Longitudinal Study of Ageing. Participants were followed from Wave 2 (2004–2005) to Wave 6 (2012–2013). The 8‐item Centre for Epidemiologic Studies Depression (CESD) scale was used to measure depressive symptoms in the past week, which were categorized as no/low, mild, or high. Normal glucose levels and prediabetes were defined using baseline haemoglobin A1c measurements. Incident diagnosed diabetes was reported by participants. Cox regression estimated hazard ratios of incident diabetes associated with depressive symptoms and prediabetes. Results A total of 157 participants were diagnosed with diabetes over a mean of 6.7 years. Relative to participants with normal glucose levels and no/low depressive symptoms at baseline, the adjusted hazard ratios were 0.85 (95% CI 0.40–1.82) and 1.62 (95% CI 0.84–3.15) for those with normal glucose levels and mild depressive symptoms and normal glucose levels and high depressive symptoms. The adjusted hazard ratios for participants with prediabetes and no/low depressive symptoms, mild depressive symptoms, and high depressive symptoms were 4.84 (95% CI 3.08–7.60), 7.17 (95% CI 4.00–12.88), and 7.77 (95% CI 4.33–13.93), respectively. Conclusions Older adults with elevated depressive symptoms and prediabetes have an increased risk of diabetes compared to those with only one of these risk factors. Copyright © 2016 John Wiley & Sons, Ltd.