The prevalence of obesity in schizophrenia patients is high, especially in chronic and medicated patients. Few studies have explored the relationships between obesity, cognition and clinical correlates in patients with schizophrenia. This study was designed to assess the prevalence and clinical correlates of obesity and its relationship to cognitive impairment in Chinese patients with schizophrenia. We recruited 633 inpatients and collected clinical, demographic data and lipid parameters. The Positive and Negative Syndrome Scale (PANSS) and its five-factor model were adopted for psychopathological symptoms. The prevalence of comorbid obesity in schizophrenia patients was 16.4%. The plasma levels of glucose, triglyceride, low density lipoprotein (LDL), apolipoprotein B, and cholesterol were higher, but high density lipoprotein (HDL) levels were lower in obese patients than those in non-obese patients (all p <0.05). Furthermore, obese patients had lower PANSS negative symptom, cognitive factor and total scores than non-obese patients (all p <0.05). Correlation analysis showed a significant correlation between BMI and the following variables: age, marriage, gender, negative symptoms, general psychopathological symptoms, cognitive factor, PANSS total score, glucose, triglycerides, HDL, LDL, cholesterol and apolipoprotein B (all p <0.05). Further multiple regression showed that PANSS cognitive factor, PANSS total score, and triglyceride were important independent predictors of obesity. Our results indicate a high prevalence of obesity in Chinese patients with chronic schizophrenia. Multiple demographics, clinical variables, and lipid parameters are associated with obesity in schizophrenia. Moreover, obesity appears to be a protective factor for psychological symptoms. However, not having objective assessments for cognition in this study is a limitation. (C) 2019 Elsevier B.V. All rights reserved.