Since 1971, a series of annual Environmental Surveillance … reports have served as the official public record of Los Alamos National Laboratory's (LANL) environmental performance. In northern New Mexico, where past LANL emissions are a public health concern, there is public skepticism over the accuracy of information contained in these reports. To test the hypothesis that LANL Environmental Surveillance … reports systematically understate past emissions, we compared the data on releases in LANL's own internal Occurrence Reports Collection (ORC) to the data reported to the public in the Environmental Surveillance … reports. A data set of 89 environmental occurrences recorded in the ORC in the time period from 1971 through 1980 was assembled. We did not find a systematic pattern of quantitative underreporting of source terms. However, 17 of the 89 (19%) environmental occurrences recorded in the ORC were not reported to the public in the Environmental Surveillance … reports. The observed discrepancies are discussed in terms of their relevance to public health concerns. Methodological caveats dictate restraint in applying these findings beyond the scope of the relative comparison performed here. Possible social origins for the rejected hypothesis are discussed. Areas for further consideration by the Centers for Disease Control's dose reconstruction study of LANL are identified.