Milk urea nitrogen (MUN) is a monitoring tool to evaluate dietary crude protein (CP) and predict urinary urea-N (UUN). However, MUN is also heritable (h² = 0.21–0.41) suggesting possible genetic effects. In this study we evaluated the interaction between dietary CP (14.9 vs. 16.5%, dry matter basis) and cow phenotypic MUN group on performance, milk protein yield, and feed conversion efficiency using 18 cows grouped in 9 pairs consisting of 1 high MUN cow (HMUNC; 15.5 ± 1.5 mg/dL) and 1 low MUN cow (LMUNC; 12.2 ± 1.3 mg/dL) of same parity, similar DIM (89 ± 18 DIM), and identical milk protein yield (1.39 ± 0.22 and 1.38 ± 0.21 kg/d for HMUNC and LMUNC, respectively). The experiment was designed as a 2-period crossover (26 d with last 3 d for sampling) with varying levels of soyhulls and soy Plus in the diets. Dry matter intake (DMI) was recorded using an “e-gate” system. Dry matter intake (27.7 ± 3.7 kg/d), fat-protein-corrected milk (FPCM; 41.6 ± 7.6 kg/d) and yield of milk protein (1.36 ± 0.26 kg/d), fat (1.5 ± 0.4 kg/d) and lactose (2.4 ± 0.5 kg/d) were not influenced by dietary CP and cow phenotypic MUN group. Reducing dietary CP reduced MUN (10.7 vs. 14.1 mg/dL; P < 0.01) and MUN yield (5.0 vs. 6.9 g/d; P < 0.01), and it tended to reduce feed conversion efficiency (kg FPCM/kg DMI; 1.14 vs. 1.18; P = 0.09). The LMUNC had lower MUN (11.4 vs. 13.4 mg/dL) and MUN yield (5.51 vs. 6.35 g/d) than HMUNC. No interactions were found between cow phenotypic MUN group and dietary CP for MUN (P = 0.88), MUN yield (P = 0.85), FPCM (P = 0.78), milk protein yield (P = 0.60) and feed conversion efficiency (P = 0.49). Although more research is needed, the lack of interactions suggested that current recommended MUN target values should apply to cows (or a herd) regardless of their potential for milk protein production. Results suggested also that genetic selection could be applied to reduce MUN (and presumably UUN excretion) and MUN yield without penalizing DMI, milk production, and yields of milk protein, fat and lactose.