Recently, the “Remission in Schizophrenia Working Group” proposed remission criteria consisting of a reduction to mild levels on key symptoms for at least 6 months.This study applied these remission criteria to a large first-episode psychosis sample in order to (1) determine the rates of remission; (2) explore predictors of remission; and (3) test the external validity of these criteria.We analyzed data from 462 subjects with a first-episode of psychosis who participated in a long-term, multinational, randomized, double-blinded trial of risperidone and haloperidol over 2 to 4 years.At some time point in the study 323 (70%) of the 462 subjects had a reduction to mild levels on the key symptoms as measured by the PANSS although only 109 (23.6%) maintained this level for at least 6 months thereby meeting remission criteria. The two strongest predictors of remission were shorter duration of untreated psychosis (p=0.01) and treatment response at 6 weeks (p=0.001). Compared to non-remitted patients, those in remission experienced greater improvement on all PANSS subscales (p<.0001), CGI-S (p<.0001), better quality of life (p=0.006), fewer relapses (p<.0001), displayed a more favorable attitude towards their medication (p=.002), had lower EPS levels according to the ESRS (p=<.0001) and received lower doses of antipsychotic medication (p=0.003). The remission and non-remission groups did not differ significantly regarding composite cognitive scores, suicidality and body mass index.The results suggest that the remission criteria, although based solely on core symptom improvement, can effectively identify patients who have a more favorable overall outcome.