Estimates of the concentration of cadmium constituting a threat to aquatic ecosystems were derived from laboratory tests conducted at two levels of the biological hierarchy. A population level estimate was derived from single species toxicity tests and a community level estimate was derived from laboratory tests on microbial communities. Estimates were compared to each other and to an ecosystem level estimate derived from reports of ecological health and ambient cadmium levels in rivers, lakes, and streams. Estimates of permissible levels for short term exposures differed by an order of magnitude. Single species toxicity tests indicated that a level of 46.1 ug Cd/L would affect only 5% of taxa. The corresponding estimate from the community level test was 459.4 ug Cd/L. Similar estimates of permissible levels for chronic exposures were not significantly different (1.02 and 0.20 ug Cd/L, single species arid community level tests, respectively). Both of the laboratory derived estimates of permissible levels for chronic exposure fell within a rational range; the minimum level defined by median cadmium levels reported in healthy aquatic systems (0.05 ug Cd/L), and the maximum level defined by median cadmium levels reported in damaged aquatic systems (9.2 ug Cd/L). However, the community level estimate was obtained more efficiently, permitting an estimate of effects on diversity from a single test. Single species level tests, community level tests, and field studies each contributed unique information to hazard evaluation. Using information from all levels will strengthen predictions. / Master of Science