Restriction of both dietary fat and extrinsic sugars is standard advice for weight reduction. It has been suggested that foods, and diets, that combine high levels of sugars and fat particularly contribute to overconsumption. Weighed dietary data on 1087 men and 1110 women aged 16-64 who took part in the Dietary and Nutritional Survey of British Adults were examined for evidence of this hypothesis. Consumption of the main sugary fatty foods (cakes, biscuits, confectionery and puddings) averaged 12% of energy for men, 14.9% for women. Intake of these foods (as a proportion of total energy) was inversely related to BMI (adjusted for age and smoking). After exclusion of the substantial number who were dieting or unwell or who may have under-reported their intakes, the association remained significant only in men. Consumption of sugary fatty foods showed a positive association with intakes of fibre, a negative association with vegetables and no relationship with percentage of energy from fat. For the investigation of diet composition, men and women were divided into four groups, high or low in extrinsic sugars energy (cut point 15%) and fat energy (cut point 40%). For men consuming high fat diets (> 40% energy) mean BMI was higher in the low sugar group. After exclusion of dieters and unwell, men with low sugar intakes still had a higher mean BMI than men with high sugar intakes. BMI and extrinsic sugars energy were still negatively but weakly correlated (r = -0.10; P < 0.05) after adjusting for age, smoking, energy, fat intake, and dieting/under-reporting. In conclusion, there is little evidence in this cross-sectional survey that either sugary fatty foods, or diets high in sugars, are associated with obesity. Rather, sugars appear to have a weak negative association with BMI that is not totally explained by confounders such as dieting, under-reporting or the inverse correlation between energy from sugars and fat.