OBJECTIVE: Different myths about schizophrenia endorsed by clinicians maintain the pessimism about outcome thus reducing chances of improvement. There are no recent North American studies on the long-term outcome of first-episode schizophrenia to clarify if these beliefs are myths or reality. Our study describes the long-term outcome (10 to 16 years) of a first-episode schizophrenia incidence cohort (n = 142) in a Canadian urban centre between 1983 and 1999. METHOD: Clinical and social functioning at different time points were assessed retrospectively from medical files of a catchment area hospital in Montreal. Service use and deaths were noted from provincial databases of physician billings, hospitalization, and vital statistics. RESULTS: Hospitalization days decreased considerably after the first year, with a small minority still needing it episodically after 4 years. Marital and occupational status were generally stable over time, but autonomy in living arrangements worsened. Thirty-three percent of subjects quit the Catchment Area Specialized Psychiatric Services. This group showed better social functioning while they were followed, were hospitalized less afterwards, and had fewer suicides, therefore indicating a better outcome for them. At the end of our study, 15% of the patients still alive were well enough to function without seeking medical help and 25% were not taking antipsychotic medication. Better outcome was predicted by older age at admission, being married, higher premorbid autonomy in living arrangements, and female sex. CONCLUSION: A significant proportion of first-episode schizophrenia patients achieve moderate long-term outcome, and the stability of global functioning is more frequent than deterioration, as shown in most industrialized countries.