Evidence has shown that soybean meal is perceived as more palatable than canola meal by dairy calves in short-term preference tests. This study evaluated the effect of protein source on longer-term dietary selection of dairy calves. In experiment 1, 40 Holstein bull calves (11.4 ± 4.3 d of age) were randomly assigned to 1 of 2 choice diets for 6 wk: base starter pellet (S; 12% crude protein; CP) and high-protein pellet (40% CP) containing either (1) soybean meal (SB) or (2) canola meal (CM). In wk 7 to 8, all calves were offered a single pelleted diet containing the protein source to which they were previously exposed. In experiment 2, 22 Holstein bull calves (9.9 ± 4.6 d of age) were offered, for 6 wk, a choice of 2 mixed pelleted diets: (1) 70% S and 30% SB (SB mix), or (2) 70% S and 30% CM (CM mix). In wk 7 to 8, calves were randomly assigned to 1 of 2 choice diets, as in experiment 1: (1) SB + S, or (2) CM + S. All feeds were provided ad libitum. Calves received 6 L/d of milk replacer [0.75 kg/d of dry matter (DM)] for the duration of both experiments. Feed intake was recorded daily and calves were weighed every 14 d. Feeds were sampled weekly to analyze DM and nutrient intake. Mixed diets in experiment 2 were analyzed for CP in wk 4 and 6 to assess feed sorting (calculated as actual CP intake as a percentage of predicted intake). In experiment 1, calves offered SB + S in wk 1 to 6 consumed more high-protein pellet than calves offered CM + S [73 vs. 42% of DM intake (DMI)] and, consequently, more CP (168 vs. 117 g/d). Solid feed DMI and average daily gain were similar between treatments. When offered a single diet in wk 7 to 8, calves offered starter containing soybean meal increased intake to a greater extent than calves offered the starter containing canola meal. In experiment 2, calves preferred the SB mix to CM mix (preference ratio: 0.7). Calves consumed more CP than predicted from SB mix in wk 4 and 6 (108 ± 2.0%), indicating that they were sorting in favor of SB. In contrast, calves consumed less CP than predicted from CM mix in wk 4 (81.48 ± 4.1%), indicating that they were sorting against CM. When assigned to choice treatments in wk 7 to 8 of experiment 2, calves offered SB + S consumed more protein pellet than calves offered CM + S (81 vs. 31% DMI) and consumed more CP (378 vs. 196 g/d). Average daily gain was greater for calves offered SB + S but DMI was similar. Overall, these results suggest that dietary selection was influenced by innate feed preferences, and milk-fed calves may not be sensitive to protein imbalances in their diet.