In an ongoing prospective investigation of the course and outcome of schizophrenia, the global functioning (employment, social contacts, symptoms) and rehospitalization rate of 72 patients were assessed one year after clinical discharge. With regard to chronicity of illness, the global outcome was better for first admissions than for multiple admissions. On controlling the patients' functioning one year before index admission, this difference did not hold true. Hence, there must be difference in the functioning of the two groups which dates back a rather long time. However, multiple admissions showed an actual downward trend in their work functioning (p less than 0.05). Rehospitalization rates of the two groups did not differ significantly (33% for first admissions versus 40% for multiple admissions). Evaluation of differences in drug-taking behavior revealed that patients with better global functioning tended to take their neuroleptics not as regularly as prescribed. Whereas for good functioning first admissions, taking into account, their better spontaneous course (rehospitalization rate 27%), an intermittent neuroleptic strategy may be in order, for multiple admissions the reversal seems to be true (rehospitalization rate for compliant patients 27%, for non-compliant patients 73%, p less than 0.01). However, a small group of good functioning multiple admissions did not have to be rehospitalized in spite of non-compliance, whereas the patients with the poorest function relapsed despite their compliance. We conclude that, whereas for first admissions long-term neuroleptic medication may not be generally indicated on account of their prognostic heterogeneity most of the multiple admissions will profit by this therapeutic strategy in respect of rehospitalization.(ABSTRACT TRUNCATED AT 250 WORDS)