Background: There is a difference between measured and predicted weight loss in obese patients. This might be explained by the composition of weight loss, adaptive thermogenesis, or poor compliance. Patients and Methods: 48 overweight and obese female patients (31.5 ± 6.1 years; BMI 35.4 ± 4.4 kg/m2) were investigated before and 13.9 ± 2.4 weeks after dietary treatment (1,000 kcal/day). Body composition was measured by air-displacement plethysmography and resting energy expenditure (REE) by indirect calorimetry. Physical activity was assessed using electronic pedometers in order to calculate total energy expenditure from REE and physical activity level (PAL). Fat mass (FM) and fat-free mass (FFM) were converted into caloric equivalents using 9.45 kcal/g FM and 1.13 kcal/g FFM. Predicted weight loss was calculated by Wishnofsky’s ‘7,700 kcal/kg rule’. Results: Weight (–8.4 ± 3.9 kg; p < 0.001), FM (–7.8 ± 3.6 kg; p < 0.001), and FFM (–0.6 ± 2.0 kg; p < 0.05) decreased with caloric restriction. Measured weight loss was only 44% of the predicted value. Since FM contributed to 87% of weight loss, the energy deficit/kg weight loss was considerably higher (9,098 ± 2,349 kcal/kg) than the assumed 7,700 kcal/kg. Adaptive thermogenesis after weight loss was significant in 26 of 48 women (–3.2 ± 1.2 kcal per kg FFM; p < 0.001). Conclusion: 14% of the difference between measured and predicted weight loss was explained by the higher proportion of FM in weight loss and 38% by adaptive thermogenesis (in 54% of the women). Thus, poor compliance was responsible for about 50% of the difference between measured and predicted weight loss only.