ELISA is the currently accepted method for screening rodent colonies for Mycoplasma pulmonis infection. While this assay has greatly improved mycoplasmal detection, it suffers from major defects. Cross-reactions with M. arthritidis are the major technical problem, and prevent definitive diagnosis. Current methods for obtaining a definitive diagnosis are accurate in about 80% of cases, and include ELISA testing for both organisms, immunoblot analysis, and blocking of the murine reaction with heterologous serum. Another technical difficulty is the inherent variability in the assay, which can be overcome by rigid quality control measures and careful attention to detail. The difficulties that arise from the natural history of mycoplasmal infection in barrier-maintained colonies, i.e., low incidence of infected animals and delayed antibody response in animals infected with low numbers of organisms, seriously limit the usefulness of the ELISA. While the assay can be extremely useful in screening breeding colonies and in eliminating mycoplasmas from such colonies, it cannot easily be used to screen potential sources of weanling animals for experimental use.