Objectives : The aims of this study were to discriminate the clinical differences, to measure the estrogen and homovanillic acid levels, to evaluate a correlation between estrogen and homovanillic acid, and to identify an association of cognitive deficit with estrogen and homovanillic acid among male and female schizophrenics. Methods : In addition to the structured interviews, the plasma estrogen levels by radioimmunoassay and the homovanillic acid levels by HPLC were measured in 20 male and 21 female schizophrenics as well as 10 healthy male and 9 female controls. Results : 1) The plasma estrogen levels were higher in females than males, and significantly higher in female schizophenics than female controls. The homovanillic acid levels were higher in female schizophrenics than female controls, and were lower in male schizophrenics than male controls. 2) The onset age seemed to be earlier in male schizophrenics, and the frequency of admission, duration of antipsychotic drug administration, dosage of antipsychotics and duration of illnesses were more in males. The estrogen and homovanillic acid levels were significantly higher in female schizophrenics. 3) The estrogen levels had a significant positive correlation with sex, age and onset age, while the homovanillic acid levels did with sex. However, estrogen were not correlated with homovanillic acid levels. 4) The estrogen and homovanillic acid levels were not significantly different between male and female schizophrenics with cognitive deficits. In the schizophrenic patients without cognitive deficits, the estrogen levels were significantly higher in females, while there were no significant sex differences in homovanillic acid. 5) In the male and female schizophrenics predominantly with negative symptoms, there were no significant differences in estrogen and homovanillic acid levels. In those predominantly with positive symptoms, the estrogen levels were significantly higher in females, while there were no sex differences in homovanillic acid levels. 6) In schizophrenics with undifferentiated subtype, the estrogen and homovanillic acid levels were significantly higher in females. In those with paranoid or disorganized subtypes, the estrogen levels were significantly higher in females, while there were no sex differences in the homovanillic acid levels. 7) The mean values of PANSS-negative, PANSS-total, PANSS-CF, MMSE-K and estrogen levels were significantly higher in male schizophrenics with cognitive deficits. The mean values of illness duration, CGI, PANSS-positive, PANSS-negative, PANSS-total, PANSS-CF and MMSE-K were significantly higher in female schizophrenics with cognitive deficits. 8) The variables which showed significant correlation with cognitive deficits were PANSS-negative, PANSS-total, PANSS-CF, MMSE-K and estrogen levels in male schizophrenics. The variables which showed significant correlation with cognitive deficits were subtypes, onset age, illness durataion, CGI, PANSS-positive, PANSS-negative, PANSS-total, PANMSS-CF and MMSE-K in female schizophrenics. The estrogen levels were significantly correlated with admission frequencies, history of antipsychotic administration, duration of antipsychotic administration and cognitive deficits in male schizophrenics, while age were not correlated with in females. The homovanillic acid levels had a significant correlation with subtypes and onset age in male schizophrenics, while there were no correlation among variables in females. Conclusions : Although the plasma concentrations of estrogen and homovanillic acid in female schizophrenics were significantly higher than males, we could not find an association between them. Furthermore, the various factors affecting on the cognitive deficits, estrogen and homovanillic acid levels seemed to be somewhat different according to sex.