We assessed the response of phytoplankton communities in aquatic microcosms to single applications of liquid creosote. The creosote was applied to 14 microcosms at concentrations ranging from 0.06 to 109 mg/L. Two microcosms served as controls. Phytoplankton samples were collected from each microcosm one week and 1 d before treatment and at 7 and 21 d after treatment. Temporal changes (response-recovery) in phytoplankton community composition were assessed with principal response curves. Creosote had no direct toxic effect on the phytoplankton community based on total abundance and number of taxa. Population levels declined in all treatments between day -1 and day 7, but this trend mirrored a similar decline in the control microcosms. At both 7 and 21 d after treatment, population densities and number of taxa in most treatments exceeded those in the controls and exhibited a parabolic relationship relative to creosote concentration. This relationship was most pronounced at 21 d, at which time population densities and number of taxa at intermediate concentrations were up to twice those at low and high concentrations. This response pattern seems to represent an indirect response to impacts on zooplankton and a corresponding reduction in grazing pressure. In contrast, total algal biomass declined 52 to 97% relative to the controls at all but the lowest creosote concentration at 7 d. This apparent decline was due to a significant proliferation of the alga Closterium moniliforme in the controls and low creosote concentration. At 21 d, no difference was found in total biomass between treated and control microcosms. The results of this study suggest that creosote does not pose a significant direct risk to phytoplankton at concentrations likely to be encountered in most contaminated aquatic environments; however, stimulation of algal populations could occur in situations of long-term chronic exposure or spill events that remove predatory zooplankton populations.