The objective of this work was to update and evaluate predictions of essential AA (EAA) outflows from the rumen. The model was constructed based on previously derived equations for ruminally undegraded (RUP), microbial (MiP) and endogenous (EndP) protein rumen outflows, and revised estimates of the EAA composition of each protein fraction. Corrections were included to account for incomplete recovery of EAA during 24-h acid hydrolysis. The predicted protein and EAA rumen outflows were evaluated against a data set of observed values from the literature. Initial evaluations indicated a mean bias for nonammonia, nonmicrobial N flow ([RUP + EndP] / 6.25) of 12 g N/d. Root mean squared errors (RMSE) of EAA predictions ranged from 29.9 to 43.7% of observed mean values. Concordance correlation coefficients (CCC) ranged from 0.29 to 0.48. Except for Leu, all EAA rumen outflows were overpredicted by 4.0 to 35 g/d. In addition, significant slope bias was present for all EAA except Met and Leu. It is unclear if the observed bias in EAA flows reflects problems in the prediction equations. Under the assumption that the mean and slope biases may be due to problems in the prediction equations, regression equations were derived to adjust predicted EAA flows. The residuals of the EAA flows were regressed on each of the 3 protein flows. Residuals were correlated with each of the 3 protein flows for all EAA. After applying these new linear regressions, RMSE for EAA predictions ranged from 25.0 to 32.9% of observed mean flows. The mean bias was removed for all predicted EAA flows; however, a small slope bias was introduced for Lys, and Thr. The CCC ranged from 0.43 to 0.57. For the purposes of predicting post-ruminal EAA flows and based on fit statistics, adjusting the mean and slope biases using the adjustments from the protein EAA flows yielded marginally better results than predicting EAA flows without adjustments. Future work should focus on identifying the cause of the observed prediction bias.