Experiments are described that have lead to the development of a highly reproducible tryptic map of recombinant DNA derived bovine somatotropin (rbSt). Tryptic digestion of rbSt at 37 degrees C results in the formation of a precipitate. Preliminary characterization of the precipitate suggests that its formation is due to the association of intermediate tryptic fragments. An examination of the temperature dependence of the digestion has revealed that precipitate formation is inhibited when digestion is performed at 10 degrees C or less. The combination of a 5-mg sample, the use of highly purified trypsin, and digestion at 5 degrees C generate a tryptic map that exhibits an average 1.3% RSD (0.5-3.6%) for all anticipated fragments. Validation studies demonstrate that while the peak response precision is rugged to daily variation of operators or chromatographic systems, the fragment retention is not. This dictates that peaks be assigned by qualitative pattern recognition. Assay ruggedness in the peak response domain allows for the implementation of quantitative methods for the comparison of rbSt reference standard and sample tryptic maps. The assay is linear for all anticipated fragments within 50-150% of the operating range. Specificity is established by assay of pituitary somatotropins from other species and rbSt analogs produced by site-specific mutagenesis. The data demonstrate that all single amino acid substitutions examined are identified by using the technique. Assay sensitivity is validated for selected tryptic fragments through analysis of reference standard digests spiked with known amounts of rbSt analog digests. The data indicate that potential impurities of 3.2, 2.0, and 4.5% can be quantitated with statistical confidence in the tryptic fragments T1, T10, and T23 + 25, respectively.