The objective of this study was to evaluate the usefulness of correcting net portal absorption (NPA) of urea-N, ammonia, and AA-N for differences in methods before their inclusion into a meta-analysis. It was hypothesized that the difference, or portal-drained viscera (PDV) balance, between N inputs (apparently digested N plus urea-N) and outputs (ammonia plus AA-N) was 0 in the absence of measurement errors and based on the assumption that other sources of N inputs and outputs were relatively small and balanced each other. A database was built from 44 publications that reported data from 129 treatments (sheep, n = 71; beef cattle, n = 32; and dairy cows, n = 26). When necessary, NPA of urea-N (n = 38) and ammonia (n = 35) results were recalculated on a whole-blood basis, whereas NPA of AA-N (n = 87) was recalculated for all the N from AA transferred across the PDV rather than only the N from the alpha-amino group. Before corrections, PDV balance averaged 22.9% of N ingested (SD 29.0) for all treatments; after corrections, PDV balance significantly decreased to 10.2% of N ingested (SD 34.7). No difference in PDV balance was observed among species before or after corrections. Correcting NPA of urea-N, ammonia, and AA-N increased the accuracy without improving precision. Therefore, from a biological perspective, recalculating reported data seems appropriate to reduce bias due to differences in methods because this approach reduces the excess in N inputs relative to N outputs.