Objective With a prevalence that varies between 20% and 65%, poststroke depression (PSD) is a frequent sequel of stroke. The aim of this study was to determine incidence and risk factors for PSD 18 months after stroke. Methods As part of the Middelheim Interdisciplinary Stroke Study, patients were followed up for 18 months in this prospective and longitudinal epidemiological study. Clinically significant signs and symptoms of PSD were quantified by means of the Cornell Scale for Depression (CSD) and the Montgomery and Åsberg Depression Rating Scale. Activities, including social activities, were measured with the Stroke Impact Scale (SIS). Relational problems since stroke onset were defined by a questionnaire. Results Data analysis was performed on 125 patients who completed follow‐up assessments. Depression (CSD score ≥8) was diagnosed in 28% of the patients. Patients with PSD were more dependent for activities of daily living and displayed more physical and cognitive impairment than patients without PSD. The risk to become depressed decreased with 5% when the patient's activities increased with one unit on the SIS (odds ratio (OR) = 0.95; 95% confidence interval (CI) = 0.93–0.97). Patients with persistent relational problems since stroke onset had approximately four and a half times greater risk of becoming depressed than patients without (OR = 4.48; 95%CI = 1.17–16.87). Conclusions Multiple regression models indicated that the most determining features for developing PSD at 18 months poststroke include reduced activity and relationship problems due to stroke. Further studies on risk factors for PSD are essential, including psychosocial aspects, given its negative impact on rehabilitation and quality of life. Copyright © 2013 John Wiley & Sons, Ltd.