To determine the effects of preloads differing in energy density and fat content on subsequent lunch intake, 24 preschool children consumed 113g ice-cream pre- loads as mid-morning snacks (containing 0,12, or 18 g of fat, and 4 g of protein), followed nearly 2 h later by an ad libitum lunch consisting of a variety of foods. A fat-free baseline preload, consisting of dry cereal and apple juice, was also included. The baseline, fat-free, medium and high-fat preloads contained 334, 740, 953, and 1150 kJ per serving. All three ice-cream preloads produced significant but equivalent suppression of ad libitum lunch intake relative to baseline; lunch intake was not related to the fat content or energy density of the ice-cream preloads. Compensation at lunch was not macronutrient specific; children did not increase their fat intake at lunch following the fat-free preload. As energy and fat content of the preloads increased, children consumed significantly fewer foods at lunch, confirming previous findings. These changes in the variety of foods consumed were related to preference: regardless of preload, children continued to consume their preferred foods, while non-preferred foods tended to be eliminated following the higher energy preloads. Because compensa- tion for energy was incomplete, energy intake of meals plus snacks increased with the energy content of the preload. Additional research is required to determine whether manipulating the fat and energy content of snacks produces similar effects when the unit of analysis is the child′s total daily diet, rather than a single meal. © 1993 Academic Press.