The value of the lecithin/sphingomyelin (L/S) ratio in the management of high-risk pregnancies has been widely accepted. A group for which major controversy still exists is the pregnancy complicated by diabetes mellitus. Some studies, such as our own, report great success with the use of the test in this disorder, while others have been unable to rely upon it. This problem prompted a closer look at our own data, with a specific analysis of the factors which may explain the discrepancies. Specifically, a subgroup of our population was selected by limiting the study to 93 patients in White Classes B through R whose L/S ratio determinations were 2.0 or above within three days of delivery. A comparison was made of this group of patients with those of the most recent conflicting report, that of Mueller-Heubach and co-workers at the Magee-Women's Hospital in Pittsburgh. The study populations appeared similar in the distribution of disease severity and gestational ages of those patients with L/S ratios over 2.0. The incidence of respiratory distress syndrome (RDS) in those infants with L/S ratios over 2.0 was compared. The group in Pittsburgh reported an 18% incidence, while we have observed a 5% incidence in this population. The factors which may explain this difference include the definition of RDS, the inclusion of Class A patients in a series, and methodologic differences in the preparation of specimens for the determination of the L/S ratio. None of these factors can conclusively explain the disparity in these studies. However, when taken together they do appear to offer a plausible explanation for such differences. It appears essential to standardize both laboratory procedures and criteria for evaluation of neonatal outcome before comparing results.