Data sets from North American (NA, 739 diets) and North European (NE, 998 diets) feeding trials with dairy cows were evaluated to investigate the effects of dietary crude protein (CP) intake and ruminal degradability on milk protein yield (MPY) and efficiency of N utilization for milk protein synthesis (MNE; milk N / N intake) in dairy cows. The NA diets were based on corn silage, alfalfa silage and hay, corn and barley grains, and soybean meal. The NE diets were based on grass silage, barley and oats grains, and soybean and rapeseed meals. Diets were evaluated for rumen-degradable and undegradable protein (RDP and RUP, respectively) concentrations according to NRC (2001). A mixed model regression analysis with random study effect was used to evaluate relationships between dietary CP concentration and degradability and MPY and MNE. In both data sets, CP intake alone predicted MPY reasonably well. Addition of CP degradability to the models slightly improved prediction. Models based on metabolizable protein (MP) intake predicted MPY better than the CP or the CP-CP degradability models. The best prediction models were based on total digestible nutrients (TDN) and CP intakes. Similar to the MPY models, inclusion of CP degradability in the CP (intake or concentration) models only slightly improved prediction of MNE in both data sets. Concentration of dietary CP was a better predictor of MNE than CP intake. Compared with the CP models, prediction of MNE was improved by inclusion of TDN intake or concentration. Milk yield alone was a poor predictor of MNE. The models developed from one data set were validated using the other data set. The MNE models based on TDN and CP intake performed well as indicated by small mean and slope bias. This meta-analysis demonstrated that CP concentration is the most important dietary factor influencing MNE. Ruminal CP degradability as predicted by NRC (2001) does not appear to be a significant factor in predicting MPY or MNE. Data also indicated that increasing milk yield will increase MNE provided that dietary CP concentration is not increased, but the effect is considerably smaller than the effect of reducing CP intake.