The toxic effects of the herbicide Atrazine (2-chloro-4-ethylamino-6-isopropylamino-s-triazine) on the three common fresh water phytoplankton species Scenedesmus quadricauda, Chlamydomonas musicola, and Chlorella vulgaris were examined. Initially the EC50 values for atrazine were determined for each species using standard batch culture methods. The EC50 values were then recalculated from growth data taken from unialgal continuous culture experiments, and a comparison was made between the results of these two bioassay protocols. The species were then grown in pairs, in continuous cultures, in competition for phosphate. From cell density data the competition coefficients ($\alpha\sb{1,2}$ and $\alpha\sb{2,1}$ were determined. The competitive cultures were exposed to concentrations of atrazine ranging from 0 to $1.0\times 10\sp{-6}$ mol/L. From these experiments it was possible to describe the pattern of changes in the competition coefficients with changes in levels of atrazine exposure for each of the test species pairs. Finally series of model equations were developed from the empirical data and a series of computer simulations were used to predict the outcomes of competition at different initial population densities and differing atrazine exposures. In all species pairs tested the alpha value dose responses showed a biphasic pattern, with a threshold which coincided with the thresholds of the physiological responses seen i the unialgal experiments. These alpha value dose responses led to competitive outcomes and species compositions that were not predicted from the single species toxicological tests. This multispecies competitive bioassay, holds the promise of predicting community level toxicological responses.