1. The availability of lysine in seven vegetable protein concentrates was determined using the slope-ratio assay with growing pigs and rats and with two chemical techniques.
2. With pigs, estimates of availability were lower using carcass compared to live-weight values, but there was no consistent effect of including food intake in the criterion of response. Availability estimates, using food conversion efficiency on a carcass basis were (proportion of total) linseed meal − 0·23, lupin-seed meal 0.74, rapeseed meal no. 1 0.97, no. 2 0.77, sunflower meal no. 1 0.59, no. 2 0.66, no. 3 0.54.
3. Availability estimates for lysine with rats were also lower using carcass compared to live-weight values. Estimates based on food conversion efficiency were higher than those based on weight gain. Availability estimates in lupin-seed meal and sunflower meals were in general agreement with the pig estimates, whilst that with linseed meal was considerably higher (0·82). Assays were not conducted with rapessed meal due to problems of food rejection with this meal.
4. The differences in available lysine were not detected by the chemical Silcock available-lysine test (Roach et al. 1967) nor by the direct fluoro-dinitrobenzene procedure (Carpenter, 1960).