Objectives: To examine the validity of existing prediction equations for basal metabolic rate (BMR) and two generated regression equations in healthy Chinese subjects and patients with chronic diseases. Subjects: One-hundred and thirty-four healthy Chinese volunteers of aged 16 ± 88 among staff working in Shatin Hospital and their relatives, and 30 elderly patients with heart disease, stroke and chronic obstructive airway disease (COAD) were recruited. Interventions: Height, weight, biceps and triceps skinfold thickness, body fat percentage, and BMR were measured in the healthy subjects and patients. Two regression equations were derived from 70 healthy Chinese subjects. Three existing equations (WHO, Liu and Jia equations) and two derived equations were then crossvalidated in 64 subjects and 30 patients. Results: For the healthy Chinese subjects, as well as patients, the BMR calculated by Liu was the closest to the measured BMR among all the equations, although there was slight underestimation for patients. Conclusion: This study con®rms that the Liu equation is the most appropriate for predicting BMR in healthy Chinese subjects, but it underestimates the BMR in those with chronic diseases. Fat-free mass is the best predictor of measured BMR. Sponsorship: Unrestricted research grant in nutrition from the Bristol-Myers Squibb Foundation. Descriptors: basal metabolic rate; prediction equation; Chinese