Aim：Many studies have reported a high prevalence of obesity in patients with schizophrenia compared to the general population. However, no nationwide survey of schizophrenic inpatients has been conducted in Japan. The aims of this study were to conduct a large‒scale survey of BMI among hospitalized patients with schizophrenia in Japan, identify the prevalence of obesity and underweight, and examine its associated factors.
 Methods：This was a cross‒sectional study. The subjects were 14,591 schizophrenic inpatients（7,645 males and 6,946 females）with complete data from 89 facilities in Japan that agreed to participate in this study. Survey items included age, gender, height, weight, BMI, length of hospital stay, antipsychotic agents, and types of hospital ward, and the survey took place between July to December 2010.
 Results：When the BMI distribution was compared between the general adult population and patients, the rate of obesity was 26.1％ in the general adult population and 21.9％ in the patients, being lower among the patients, and the rate of underweight was 7.0％ in the general adult population and 19.5％ in the patients, being 2.8 times higher among the patients. Subsequently, logistic regression was used to examine factors associated with obesity and underweight. As the results, the obesity odds ratio was 1.32 times higher in patients in their 30s than in their 20s, 1.31 times higher in females than males, 0.62 times higher with a≥1‒year hospital stay than＜3‒month stay, 1.42 times higher in patients receiving≥4 antipsychotic agents than those receiving single typical agents, and 0.58 times higher in patients kept in a locked ward than those in an open ward. The underweight odds ratio was 2.03 times higher in patients in their≥70s than in their 20s, 1.18 times higher in females than males, 1.47 times higher with a 3‒month＜1‒year hospital stay than ＜3‒month stay, 0.7 times higher in patients receiving≥4 antipsychotic agents than those receiving single typical or atypical agents, and 1.81 times higher in patients kept in a locked ward than those in an open ward.
 Conclusions：Obesity in Japanese schizophrenic inpatients was associated with＜3‒month hospitalization, antipsychotic polypharmacy, and an open ward, whereas underweight was associated with≥3‒month hospitalization, the administration of single typical or atypical antipsychotic agents, and a locked ward, independent from age and gender.