Historically, lactose has been reported as a calculated as residual difference of solids minus fat, protein, and a constant ash. This method generally over estimates milk lactose content by 0.2 to 0.3% lactose. Our objective was to document the within and between laboratory performance of the spectrophotometric enzymatic method (Association of Official Analytical Chemists method 2006.06) for determination of the lactose content of milk used in the USDA Federal Milk Market Order laboratories. Monthly, from January through December 2017, 6 or 7 laboratories tested 14 milk samples in duplicate and reported anhydrous lactose results. The mean relative repeatability (within lab) standard deviation (RSD_r) was 0.198 + 0.021% and relative reproducibility (between lab) standard deviation (RSD_R) was 0.352 + 0.053% for anhydrous lactose measurement in 2017 at a mean anhydrous lactose concentration of about 4.55% (mass/mass). The analytical performance of the lactose analysis method was comparable to Kjeldahl true protein analysis of milk with a mean RSD_r of 0.161 + 0.014% and RSD_R of 0.470 + 0.062% for Kjeldahl true protein measurement in 2017 at a mean true protein concentration of about 3.15% (mass/mass). To achieve better accuracy of analytical results for milk payment, dairy plant accounting for milk components, and calculation of more accurate milk energy output by dairy cows for nutrition management, the USDA Federal Milk Market Orders started reporting lactose content of milk on an anhydrous mass/mass basis starting in January 2017 and will no longer report lactose by difference.