- Introduction Post-marketing data on weight-loss medications in free living population are a necessary adjunct to data from clinical trials. Materials and methods We conducted a population-based analysis of first-time medication users based on HMO pharmacy purchasing data serving > one million adults. Results During 5 years, usage of orlistat and sibutramine more than doubled and rates were higher during the months May–Aug. As compared to non-users (n = 1,038,828), annual weight-loss drug users (n = 7175) had higher women proportion, body-mass-index (BMI), bariatric surgery history, and usage of diabetes, depression, and cardiovascular medications (p < 0.001 for all). Among users, men had higher BMI (34.4 kg/m2 vs. 32.5 kg/m2), prevalence of diabetes (25.4% vs. 10.7%) and heart disease (14.2% vs. 3.5%) than women. Mean duration of purchasing weight-loss medications was 2.1 months for orlistat and 2.9 months for sibutramine. Fewer than 2% completed 12 months of weight-loss medication therapy. Among the 25% who continued to purchase at least 4 months, BMI (sub-group analysis) reduced from 33.02 kg/m2 to 32.04 kg/m2 (p < 0.001). In a multivariate model, long-term adherence (≥4 months) to weight-loss medications was associated with use of sibutramine vs. orlistat (OR = 2.08; 95%CI: 1.76–2.45), and prevalence of diabetes (OR = 1.20; 95%CI: 1.01–1.25). Age, gender, and baseline BMI were not associated with long-term adherence. Conclusions Usage of weight-loss drugs is higher among diabetes patients. However, the poor adherence to therapy is substantially below levels reported in clinical trials.