Reports describing polyacrylamide gel electrophoresis patterns of bovine hydrophobic surfactant proteins are not consistent. In this study, we found unusual staining characteristics of these proteins that may explain some of these inconsistencies. Low molecular weight surfactant proteins extracted from bronchoalveolar lavage with organic solvent are partially delipidated with Sephadex LH-20 chromatography using chloroform and methanol. Fractions from the first protein peak are dried under nitrogen then subjected to SDS electrophoresis on 20% polyacrylamide gels. Under nonreducing conditions, silver staining identifies 5- and 26-kDa bands, and Coomassie blue identifies 6-, 12-, and 26-kDa bands. When gels are stained with Coomassie blue then silver, the 5- and 26-kDa bands stain with silver and 6- and 12-kDa bands remain stained with Coomassie blue. If gels are first stained with silver then Coomassie blue, similar results occur. We modified the silver staining protocol by treating gels with dithiothreitol or 2-mercaptoethanol after electrophoresis. With this modification, 5-, 6-, 12-, 26-, and also 17-kDa bands are identifiable. Using the modified protocol and restaining gels previously stained with silver, 6-, 12-, and 17-kDa bands that were not identified previously all became visible. In further experiments, protein bands of 6-, 12-, and 26-kDa that were identified by Coomassie blue were electroeluted under nonreducing conditions. After electrophoresis of the eluted 26-kDa protein, bands of 17-, and 26-kDa under non-reducing, and 8-kDa only under reducing conditions, were apparent by using the modified silver protocol. Electroeluted 6- and 12-kDa proteins each form both 6- and 12-kDa bands with repeat electrophoresis, regardless of reducing or nonreducing conditions. Routine silver staining does not identify 6- or 12-kDa bands in these experiments. Results indicate that 8-, 17-, and 26-kDa bands are from one protein (SP-B) and polymerize readily. The 6- and 12-kDa bands are another protein (SP-C) that is resistant to reducing agents and not seen by routine silver stain. We recommend modified silver protocol for analysis of low molecular weight surfactant proteins. © 1990.