A meta-analysis of published studies was used to investigate the effect of rumen-protected methionine (RPM) added to the diets of lactating dairy cattle on dry matter intake, milk production, true milk protein (TMP) production, and milk fat yield. Differences in responses between 2 commonly used RPM products, Mepron (Evonik Industries, Hanau, Germany) and Smartamine (Adisseo, Antony, France), were investigated as well as dietary and animal factors that could influence responses. Diets were coded with respect to the amino acid (AA) deficiency of the control diet as predicted by the AminoCow model (version 3.5.2, http://www.makemilknotmanure.com/aminocow.php; 0=no AA deficiency, 1=Met deficiency, 2=Met and Lys deficiency, 3=Met and Lys plus at least 1 other AA deficiency) to test the effect of AA deficiencies on RPM response. Thirty-five studies were identified, 17 studies evaluating Mepron, 18 studies evaluating Smartamine, and 1 study evaluating both. This permitted 75 dietary comparisons between control and RPM-added diets. Diets were entered into the AminoCow and the 2001 National Research Council models to compare predictions of Met, Lys, and metabolizable protein (MP) flow. Mean Met and Lys in diets where RPM was fed were estimated to be 2.35 and 6.33% of MP, respectively. Predictions of flows between models were similar. Overall, RPM addition to diets increased production of TMP, both as percentage (0.07%) and yield (27 g/d). Dry matter intake and milk fat percentage were slightly decreased, whereas milk production was slightly increased. Differences between products were detected for all production variables, with Mepron-fed cows producing less TMP percentage but greater milk production, resulting in twice as much TMP yield. Milk protein response to RPM was not related to predicted AA deficiency, calculated Met deficiency, or Met as a percentage of MP. Other dietary factors, including Lys flow (g/d), Lys as percentage of MP, neutral detergent fiber percentage, crude protein percentage, or energy balance, had no detectable effects on TMP response. When cows with a predicted positive AA balance were fed RPM, milk production increased, but when AA balance was negative, milk production decreased. Amount of RPM added to the diet was not correlated to TMP response. This study does not support the necessity of a high Lys level as a prerequisite to obtaining a TMP response to feeding RPM or the MP requirement suggested by the National Research Council model (2001). However, more dose-response studies over a wide range of milk production and dietary regimens will be required to more clearly establish AA requirements and to predict responses to RPM supplementation.