The objective of the study was to analyse 10-year mortality among persons with schizophrenia from an epidemiological perspective. This cohort study included all persons with schizophrenia (n=255) living in the northern catchment area in Uppsala in 1991, and 1275 subjects from the national population register matched for sex, age and living area. The prevalence of schizophrenia was 0.37% and the mortality rate for individuals with schizophrenia was higher than for referents: 23.0% vs. 11.2%. The higher mortality among those with schizophrenia was mainly the result of unnatural causes and cardiovascular disease, especially in men. Excess mortality from cardiovascular disease was more pronounced in middle age, irrespective of gender. Multivariate analysis revealed higher mortality among individuals with schizophrenia living in the city than among those living in less urbanized areas. People with schizophrenia die more often than those without schizophrenia from unnatural causes or circulatory diseases. Individuals with schizophrenia die sooner from circulatory diseases than those without schizophrenia. Having schizophrenia and living in the city also results in higher mortality than having schizophrenia and living in other areas. The risk of early death from circulatory disease needs to be studied in more detail to reveal the potential respective contributions of intrinsic patient vulnerability, lifestyle factors and side-effects from psychotropic drugs.