The aim of this article was to study mortality following a first-ever cerebral infarct, accounting for ischemic stroke subtypes (lacunar, cardioembolic, atherothrombotic) and relevant prognostic variables. This study was done from s a hospital-based prospective registry of all patients with a first cerebral infarct, with a high case ascertainment of first and recurrent stroke by CT. We used a cross-sectional follow-up, using standardized methods. Analyses were performed using crude comparison of mortality data and death causes between stroke subtypes. We analyzed 30-day case fatality and 1-year mortality in 30-day survivors by means of logistic regression analysis, and mortality in 1-year survivors by means of Cox proportional hazard modeling. We also constructed Kaplan-Meier survival curves, and used log-rank testing for differences between stroke subtypes. Thirty-day case fatality was 10%, 1-year mortality 15%, and after 1-year mortality 16%. Mean follow-up was 691, SD 521 days. At the end of follow-up 36% of all patients had died. Mortality was at all three time points lowest in lacunar stroke (2, 12, and 14%, respectively), intermediate in atherothrombotic stroke (10, 16, and 15%, respectively), and highest in cardioembolic stroke (23, 22, and 21%, respectively). Death related to recurrent stroke was similar in all three stroke subtypes (13-16%). Although 30-day case fatality rate was low in lacunar stroke, a quarter of lacunar stroke patients had died at the end of follow-up. Diabetes mellitus, age, stroke subtype, and initial stroke severity were independent predictors of 30-day case fatality, but only diabetes and age were consistent independent predictors for later mortality. Recurrent stroke and heart failure were important death causes. Prognosis for (future) death following a first cerebral infarct differs between stroke subtypes; lacunar stroke patients have the lowest mortality. However, lacunar stroke cannot be regarded as a mild stroke type, as after 2 years more than a quarter of such stroke patients had died. Cardioembolic stroke patients have the grimmest prognosis: more than half of them had died within 1.5 years. Better prognosis for long-term survival following stroke may be achieved by therapies which lower the risk of stroke recurrence, provide better treatment of heart failure, or both.