Shrinking margins in the dairy industry dictate a need for increasing efficiencies of feed conversion to milk and limiting feed waste on the farm. The aim was to determine if performance and efficiency were similar when feeding 2 protein sources (canola meal, CM; soybean meal, SBM) at 2 feeding levels (limit-fed, LF; ad libitum-fed, AL). Fifty-two multiparous Holstein cows (mean ± SD; 98 ± 38 DIM; 2.73 ± 1.03 parity) received a common diet during a 4-wk covariate (COV) period. Animals were sorted by residual feed intake (RFI = DMI_Act – DMI_Pred), grouped into quartiles, and then blocked to standardize MY, milk components, BW, DMI and DIM. Each cow was randomized to receive one of 4 treatments. An average of COV wk 3–4 ad libitum intake was reduced by 5% in LF diets. The amount offered was evaluated weekly and reduced in pairs if a LF cow had orts ≥0.907 kg as-fed. Average final restriction of the LF group was 92.8% of COV wk 3–4 ad libitum intake. Diets were formulated to contain 42, 18, and 40% DM of corn silage, alfalfa haylage, and concentrate mix, respectively. The COV diet was formulated to contain equal CP from SBM and CM. Statistical analyses were performed as repeated measures using the MIXED procedure of SAS. Dry matter intake (DMI) was greater for AL vs LF cows (mean ± SEM; 29.0 vs 27.7 ± 0.22 kg/d; P < 0.02) during wk 1–4. DMI was greater for CM-fed cows in wk 1 by 0.65 ± 0.22 kg/d (P < 0.04) and for SBM-fed cows in wk 4 by 0.71 ± 0.22 kg/d (P < 0.03). Protein source did not have an effect on MY. Cows fed AL had greater MY, in wk 3–4 (52.4 vs 50.4 ± 0.55 kg/d; P < 0.02). Milk fat concentration was less for LF compared with AL cows in wk 3–5 (3.63 vs 3.87 ± 0.07%; P < 0.05). MUN was lower in CM-fed cows (10.9 vs 11.6 ± 0.24; P < 0.01). Feed efficiency (FE = energy corrected milk/DMI) was greater in the LF treatment during wk 1 (1.99 vs. 1.92 ± 0.02; P < 0.02) but not different in other weeks. While the LF method was effective at equalizing DMI in SBM- and CM-fed animals, MY decreased and FE did not commensurately compensate. Continued investigation into limit feeding strategies is needed to evaluate the effect on digestion and efficiency.