Summary — Two feeding trials were carried out on dairy cows for the first 12 weeks of lactation. Diets were based on grass silage and ensiled beet pulps and distinguished by the nature of concentrates: rich in starch (S), starch + lipids (SL), and fibre + lipids (FLi ) for trial 1; fibre (F) and fibre + lipids (FL2) in trial 2. The lipid supplement used was calcium salts of palm oil fatty acids. Lipid-supplemented diets contained ca 6% fatty acids in dry matter. Results were analysed for the first 6 weeks of lactation (period 1) then for the following 6 weeks (period 2). Whatever the basal diet, calcium salt supply did not modify milk yield (30.7 and 29.6 kg on average for trials 1 and 2, respectively). In trial 1, lipid supply did not modify protein content during weeks 1-6 but decreased it during weeks 7-12 (31.5, 32.1 and 31.4 g/kg during weeks 1-6, 28.8, 27.2 and 26.4 g/kg during weeks 7-12 for diets S, SL and FL1, respectively). Although non-significant, the same trends were observed in trial 2 (31.4 and 30.7 g/kg during weeks 1-6, 29.4 and 27.5 g/kg during weeks 7-12 for diets F and FL2, respectively). In trial 1, during weeks 7-12, lactose content was lower for diet FL1 than for diet SL (48.0 vs 49.4 g/kg). For the 2 trials and the 2 periods, butterfat content did not vary (45.2, 47.4, 45.2, 44.6, 47.0 g/kg during weeks 1-6, 40.5, 40.4, 38.2, 38.1, 39.0 g/kg during weeks 7-12 for diets S, SL, FL1, F and FL2, respectively). Lipid supply decreased the proportion of short-and medium-chain fatty acids (6-14 carbons) in both trials and both periods, increased C16:0 in both trials during weeks 1-6 only, and did not modify C18:0 and increased C18:1 during the 2 periods in trial 1 and during weeks 7-12 in trial 2. No difference in proportion of these fatty acids was observed between diets SL and FLi.