An analytical method has been developed to measure the locoweed toxin, swainsonine, in locoweed plant material. Dry ground plant samples were extracted using a small-scale liquid/liquid extraction procedure followed by isolation of the swainsonine by solid phase extraction with a cation-exchange resin. Detection and quantitation of the swainsonine were accomplished using reversed phase high-performance liquid chromatography coupled to atmospheric pressure chemical ionization tandem mass spectrometry (LC-MS(2)). The limit of quantitation was estimated to be 0.001% swainsonine by weight in dry plant material, which corresponds to the lower threshold for toxicity of locoweeds. The method of analysis was applied to the analysis of Oxytropis sericea (white locoweed) and Oxytropis lambertii (Lambert locoweed) plant samples to measure the variability of individual plant swainsonine levels within populations and within species. Individual plant variability was found to be highly significant for both O. sericea and O. lambertii populations. The combined three-year mean swainsonine values taken from three populations of O. sericea ranged from 0.046% in Utah to 0.097% in a New Mexico population. Sixteen individual populations of O. lambertii were sampled from eight different U.S. states. Swainsonine was detected at levels >0.001% in only 5 of the 16 collection sites. Those populations of O. lambertii found to contain higher swainsonine levels were restricted to the most southern and western portion of its distribution, and all were identified as belonging to var. bigelovii, whereas var. articulata and var. lambertii samples contained swainsonine at levels <0.001%.