OBJECTIVES: First, to estimate the rate and characteristic of low back pain (LBP) among Hong Kong middle-aged women. Second, to explore the relationship between being overweight, having tall stature and LBP prevalence. DESIGN: A case ± control study of Hong Kong community-based middle-aged women was conducted. SUBJECTS: The study subjects were either from the University Family Medicine Clinic or from a population-based cross-sectional study on health in middle-aged women conducted by the Department of Community and Family Medicine (based on a randomized sample of telephone numbers from the Hong Kong residential telephone directory in 1991). Among the 928 potential eligible subjects from both sources, a total of 182 cases and 235 controls participated in this study. MEASUREMENTS: Data were collected from personal interviews and anthropometric measurements. Participants were asked about exclusion criteria, the occurrence of LBP and demographics factors. The anthropometric measurements included measures of body height, body weight, waist and hip circumferences. RESULTS: Among the 182 cases who entered this study, 83 women (45.6%) had 1 ±`14 days of LBP, and 99 women (54.4%) had at least 14 days of LBP in the previous 12 months. Having tall stature and being overweight were found to have no association with LBP prevalence. Conversely, high waist-to-hip ratio ( 0.8) was noted to be inversely associated with the risk of severe (at least 14 days) LBP (adjusted odds ratio 0.43, 95% con®dence interval 0.26 ± 0.70) even when controlling for other risk factors together, but not for the overall (at least one day) LBP. CONCLUSION: This case ± control study found no association between excessive weight, tall stature and an increased risk of LBP prevalence in Hong Kong Chinese middle-aged women. Conversely, the results indicate that a high waist to hip ratio was associated with a lower risk of severe LBP. Prospective studies are needed to further evaluate the association between underweightaobesity, bone mass density and LBP.