Two respiratory chamber experiments were conducted with dairy cows to compare metabolizable energy and protein utilization when feeding white or red clover with ryegrass. In experiment 1, fresh ryegrass was mixed with fresh white (WF) or red clover (RF) (60/40, on dry matter (DM) basis). Experiment 2 involved similar mixed diets in ensiled form (WS and RS, respectively), and two ryegrass silage diets, without (GS) or with supplementary maize gluten (GS+). Barley was supplemented according to requirements for milk production. Voluntary forage DM intake remained unaffected in experiment 1 and was higher (P<0·01) in experiment 2 for WS than for GS and GS+(128 v. 98 and 106 g/kg M0·75). Within experiments, no treatment effects occurred for apparent nutrient digestibilities, milk yield, and composition. Protein utilization (milk-N/N-intake) was numerically lower on all clover-based diets (0·24 to 0·25) versus GS (0·29). With added maize gluten (GS+), protein utilization decreased to 0·23, indicating that ryegrass silage (plus barley) alone provided sufficient metabolizable protein. Consequently, higher (P<0·01) urinary energy losses occurred in GS+ compared with GS, despite similar metabolizable energy intakes, and a trend for the highest plasma urea levels was found for GS+ cows (7·59 mmol/l; P<0·1). Overall, this study illustrates that the white and red clovers investigated were equivalent in energy and protein supply, also in comparison to the ryegrass. It remains open whether these forage legumes, when supplemented to a moderate-protein ryegrass, would have contributed to metabolizable protein supply or would have merely increased metabolic nitrogen load.