Seventy-two 8-week-old ram lambs from three breeds, Suffolk, Soay and Friesland, were offered one of four diets based on dried grass and formulated to have a similar fatty acid content (60 g/kg DM) and containing: Megalac (high in 16 : 0, control; Volac Ltd, Royston, Herts., UK), whole linseed (18 : 3n-3), fish oil (20 : 5n-3 and 22 : 6n-3) or whole linseed plus fish oil. The lambs were slaughtered at approximately half of their mature live weight (43, 21 and 43 kg for Suffolk, Soay and Friesland lambs, respectively). Fish oil reduced DM intake and lamb live-weight gain (P<0.001), while DM intake, live-weight gain and subcutaneous fat content were highest in Suffolk and lowest in Soay lambs. Linseed feeding doubled the proportion (x100) of 18 : 3n-3 in the longissimus dorsi from 1.4 to 3.1 and in the subcutaneous adipose tissue from 1.2 to 2.6 (P<0.001). Suffolk and particularly Soay lambs contained higher proportions of 18 : 3n-3 than Friesland lambs in the longissimus dorsi, while in the adipose tissue, Suffolk lambs had the highest level. Feeding fish oil increased the muscle proportion (x100) of 20 : 5n-3 from 0.7 to 2.3 and 22 : 6n-3 from 0.3 to 0.8 (P<0.001). By contrast, the proportions of the longer-chain n-3 polyunsaturated fatty acids were similar across all three breeds. All three lipid supplements containing n-3 polyunsaturated fatty acids increased the content of muscle trans-18 : 1 relative to the control values, but conjugated linoleic acid (cis-9,trans-11-18 : 2) only increased in the muscle of lambs fed linseed. Feeding linseed or fish oil lowered the n-6 : n-3 ratio in sheep meat, but neither diet nor breed had much effect on the polyunsaturated fatty acid: saturated fatty acid ratio.