The concepts of partial recovery and remission have become increasingly important for the evaluation of the effectiveness of schizophrenia therapeutics. The relationship of baseline symptoms and changes in symptoms to remission of psychosis was evaluated. Fifty-six outpatients with residual schizophrenia completed a double-blind trial of olanzapine versus haloperidol and were then enrolled into a one-year open-label trial of olanzapine. Out of these 56 subjects, 13 (23%) met remission criteria at the beginning of the open-label treatment and were excluded. During the one-year study, 7/43 (16%) subjects met remission criteria. These subjects had significantly lower baseline ratings for tardive dyskinesia (TD) than subjects who did not achieve remission (1.8 +/- 1.5 vs. 4.2 +/- 4.6, P = 0.03). As expected, remitted subjects had significantly greater improvements in Brief Psychiatric Rating Scale total scores, positive subscale scores and scale for the Assessment of Negative Symptoms total scores. Remitted subjects also experienced a significantly greater improvement in depressive symptoms (P = 0.001), activation (P = 0.005), and Clinical Global Impressions scores (P < 0.001), as well as greater improvements in extrapyramidal symptoms (P = 0.007) and TD (P < 0.001). These results suggest that the relationship of depressive symptoms and improved side effects to the construct of remission in schizophrenia may deserve special attention. Future studies should aim to relate remission criteria to functional outcomes, cognition, and other important symptom domains.