A GLC procedure was developed for the evaluation of diazepam, chlordiazepoxide, and flurazepam formulations to USP-NF specifications for drug content, content uniformity, impurities, and identity by retention times and peak areas. The polyimide column, instrument zone temperatures, gas flows, internal standard solution, extraction solvent, and auxiliary equipment were the same for each drug. No derivatization of the samples was required. The GLC assay values (mean of 10 individual dosage units) for diazepam and flurazepam products were in good agreement with the results obtained by the pharmacopeial composite assays. With chlordiazepoxide capsules, when the levels of the two pharmacopeial impurities determined by GLC were added to the GLC assay results (mean of 10), athe aggregate values were consistent wit the drug content results found by the nonspecific USP method. The procedure can be made sensitive to impurity levels of approximately 0.01% for 2-amino-5-chlorobenzophenone and to approximately 0.2% for 7-chloro-1,3-dihydro-5-phenyl-2H-1,4-benzodiazepin-2-one 4-oxide. With the equipment used, the estimated potential outputs in lots per working day for a complete quality profile (drug content, content uniformity, purity, and identity) were seven for chlordiazepoxide if no impurity test was required, five if such a test was required, eight for diazepam, and seven for flurazepam.