A protocol for selection of donors of milk samples is presented. Various methods for the sampling of milk are critically reviewed and a tested procedure is described. The measurement of milk volume is reviewed. Analyses for milk lipids are presented and discussed. Results of the analyses relative to various factors affecting fat content are given. Depending upon the type of study, sampling of milk by the Egnell or similar pump is recommended. For field studies, the creamatocrit procedure for estimation of fat content is acceptable if calibrated. Satisfactory research methods are the Roese-Gottlieb, modified Folch, and dry column extractions followed by gravimetry.