The A risk allele of rs9939609 of the fat mass- and obesity-associated gene (FTO) increases body fat mass. To examine whether FTO rs9939609 affects obese individuals' response to a high-fat, low-carbohydrate (CHO) (HF) or low-fat, high-CHO (LF), hypo-energetic diet and whether the effect of the FTO variant depends on dietary fat and CHO content. In a 10-week, European, multi-centre dietary intervention study 771 obese women and men were randomized to either LF (20-25% of energy (%E) from fat, 60-65%E from CHO) or HF (40-45%E from fat, 40-45%E from CHO), hypo-energetic diet (measured resting metabolic rate multiplied by 1.3-600 kcal day(-1)). Body weight, fat mass (FM), fat-free mass (FFM), waist circumference (WC), resting energy expenditure (REE), fasting fat oxidation as % of REE (FatOx), insulin release (HOMA-beta) and a surrogate measure of insulin resistance (HOMA-IR) were measured at baseline and after the intervention. In all, 764 individuals were genotyped for FTO rs9939609. For A-allele carriers the drop-out rate was higher on HF than LF diet (in AT, P=0.002; in AT/AA combined, P=0.003). Among those individuals completing the intervention, we found no effect of FTO rs9939609 genotype on Deltaweight, DeltaFM, DeltaFFM, DeltaWC or DeltaFatOx. However, participants with TT had a smaller reduction in REE on LF than on HF diet (75 kcal/24 h; interaction: P=0.0055). These individuals also showed the greatest reduction in HOMA-beta and HOMA-IR (interaction: P=0.0083 and P=0.047). The FTO rs9939609 may interact with the macronutrient composition in weight loss diets in various ways; carriers of the A allele on LF diet appear to have a lower risk for drop out, and TT individuals have a smaller decrease in REE and greater decrease in HOMA-beta and HOMA-IR on LF than on HF diet.