Individual differences in appetite are increasingly appreciated. However, the individual day-to-day reliability of appetite measurement is currently uncharacterised. This study aimed to assess the reliability of appetite following ingestion of mixed-macronutrient liquid meals at a group- and individual-level. Two experiments were conducted with identical protocols other than meal energy content. During each experiment, 10 non-obese males completed four experimental trials constituting high- and low-energy trials, each performed twice. Experiment one employed 579 kJ (138 kcal) and 1776 kJ (424 kcal) liquid meals. Experiment two employed 828 (198 kcal) and 4188 kJ (1001 kcal) liquid meals. Visual analogue scales were administered to assess appetite for 60 min post-ingestion. The typical error (standard error of measurement) of appetite area under the curve was 6.2 mm60 min-1 (95%CI 4.3 to 11.3 mm60 min-1), 6.5 mm (95%CI 4.5 to 11.9 mm60 min-1), 7.1 mm60 min-1 (95%CI 4.9 to 12.9 mm60 min-1) and 6.5 mm60 min-1 (95%CI 4.5 to 11.8 mm60 min-1) with the 579, 828, 1776 and 4188 kJ meals, respectively. A systematic bias between first and second exposure was detected for all but the 4188 kJ meal. The change in appetite with high- vs. low-energy meals did not differ at a group level between first and second exposure (mean difference: -0.97 mm60 min-1; 95%CI -6.48 to 4.53 mm60 min-1), however, ~50% of individuals differed in their response with first vs second exposure by more than the typical error. Appetite responses are more reliable when liquid meals contain a higher- vs lower-energy content. Appetite suppression with high- vs low-energy meals is reproducible at the group- but not individual-level, suggesting that multiple exposures to an intervention are required to understand true individual differences in appetite.