Variations in testing that could lead to the rejection of a material fully conforming to specification requirements, or the acceptance of a material with undesirable chemical or physical properties, are apparent in the results reported by laboratories participating in comparative tests of Portland cements.Many of the causes for variation in chemical analyses and physical test results are listed in this discussion, and remedies for some of the more frequently encountered deficiences in apparatus and methods are suggested. Particular consideration is given to problems which do not seem to have been covered in sufficient detail in previous discussions of cement testing procedures. Literature references are given for additional information. This publication supersedes NBS Monograph 28.