An experiment was carried out in Northern Ireland in 1994 to examine the effects of ensiling grass on the partitioning of nutrients between lean and fat deposition in lambs using the comparative slaughter technique. The six treatments consisted of: (1) unsupplemented grass silage; (2) grass silage supplemented with concentrates, concentrates constituting 250 g/kg total dry matter intake (DMI); (3) grass silage supplemented with concentrates, concentrates constituting 500 g/kg total DMI; (4) unsupplemented grass which had been conserved by freezing; (5) pre-frozen grass supplemented with concentrates, concentrates constituting 250 g/kg total DMI, and (6) pre-frozen grass supplemented with concentrates, concentrates constituting 500 g/kg total DMI. A total of 84 Dutch Texel×Greyface (Border Leicester×Blackface) lambs, consisting of 42 males which were initially 35 (s.d. 4·8) kg liveweight and 42 females which were 33 (S.D. 2·6) kg liveweight, were used. The silage was well preserved as indicated by its low pH (3·8), low concentration of ammonia (60 g/kg total nitrogen) and high concentration of lactic acid (129 g/kg DM). The grass had a higher pH (5·2) and higher WSC concentration (129 g/kg DM) and lower concentrations of lactic (33 g/kg DM) and acetic acids (6 g/kg DM). Ensiling had no significant effect (P>0·05) on empty body and carcass weight gains. Animals offered unsupplemented silage had significantly lower protein (P<0·001) concentration in the carcass gain and significantly higher lipid and energy (P<0·001) concentrations than animals offered unsupplemented frozen grass. However, there was no effect of forage type on chemical composition of carcass gains when 50% forage was offered. It is concluded that ensiling had no effect on empty body or carcass weight gains, but did increase carcass fatness when unsupplemented forages were offered.