To fill the feed deficit in the autumn/late lactation period in a seasonal grazing system, supplementation is required. This study aimed to investigate the use of baled grass silage or concentrate as supplementation to grazing dairy cows in late lactation. Eighty-four grass-based spring-calving dairy cows, averaging 212d in milk, were allocated to 1 of 6 treatments [high grass allowance (HG), low grass allowance (LG), grass with a low concentrate allocation (GCL), grass with a low grass silage allocation (GSL), grass with a high concentrate allocation (GCH), and grass with a high grass silage allocation (GSH)] to measure the effects of using baled grass silage or concentrate as supplements to grazed grass. Effects on intake, milk yield, milk composition and N fractions, and N utilization efficiency were measured. Treatments HG and LG received 17 and 14kg of dry matter (DM) grass/cow per d, respectively. Treatments GCL and GSL were offered 14kg of DM grass/cow per d and 3kg of DM of supplementation/cow per d. Treatments GCH and GSH were offered 11kg of DM grass/cow per d and 6kg of DM of supplementation/cow per d. Milk yield was greatest in the GCH treatment and milk solids yield was greatest in both concentrate-supplemented treatments. The HG and LG treatments excreted a greater quantity of N as a proportion of N intake than the supplemented treatments. The HG treatment also excreted the greatest total quantity of N. This indicates an improvement in N utilization efficiency when supplementation is offered compared with grazing only. Offering 6kg of DM of either grass silage or concentrate as supplementation decreased milk true protein concentration compared with offering a grass-only diet. This suggests that increasing the proportion of supplementation relative to grass may negatively affect milk processability, which is associated with milk true protein concentration.