Short-term feed preferences were studied in individually caged chickens fed sequentially in order to understand a previously described imbalance in the intake of diets offered. Sequential feeding (SF) was carried out for four 48 h cycles in male broiler chickens. The diets varied in energy (2800 (E-) and 3200 kcal/kg (E+)) and protein (230 (P+) and 150 g/kg (P-)) contents. SF was compared to standard feeding (C) (3000 kcal/kg ME and CP = 190 g/kg). In experiment 1, three treatments were used., C, SE (E- followed by E+) and SE, (E+ followed by E-). Four treatments were used in experiment 2: C, S-P (P+ followed by P-), S-E and S-EP (P+E- followed by P-E+). Total feed intake was measured during the SF period. After this, short-term preferences were evaluated with a choice test on chickens previously fed with the same feeds during the SF period (experienced birds) and in C chickens (naive birds). In both experiments, total feed intake was similar among treatments and the percentage of each feed consumed was not significantly different from controls (50%). In experiment 1, S-E and S-E' chickens over-consumed E+ and under-consumed E- diets only during the first 15 min of the fourth cycle. The choice test indicated that experienced chickens preferred E+, while naive chickens preferred E-. Similarly, in experiment 2, chickens over-consumed E+ and E+P- during the first 15 min of the fourth cycle, but the intake of diets varying in protein content was not different from controls. During the choice test, as in experiment 1, experienced chickens preferred E+, while naive chickens preferred E-. There was a slight preference for the protein-poor diet in naive birds and there was no preference in the diet varying in both protein and energy contents. Experience modified choice between feeds varying in energy content but not in protein. When feeds were known, preference for energy affected the feed intake immediately after switching from one diet to the other although lower with the diet also varying in protein, it did not influence the total intake of each diet. Interactions between the nutritional properties and sensorial cues of feed could explain these results.