The results of single species toxicity tests were compared to tests at higher levels of biological organisation. Laboratory single species acute and chronic toxicity tests using three freshwater copepod species: Calamoecia lucasi, Boeckella delicata, and Mesocyclops cf. leuckarti, were compared to community-level responses in naturally derived laboratory microcosms and in situ mesocosms dosed with pentachlorophenol (PCP). Acute 48 h lethality tests used two copepod life stages (nauplii and adults). Tests were conducted at 22°C with laboratory cultured animals of all species and at varying temperatures with seasonally collected C. lucasi adults. Chronic sublethal tests with <24 h nauplii measured time to metamorphosis. Copepod cultures were established using animals collected from Lake Rotomanuka, New Zealand. A 12-day laboratory microcosm experiment was conducted during the summer of 1997 using naturally derived microcosms (1.6 L) established using water and planktonic organisms collected from Lake Rotomanuka. The microcosm experiment coincided with the summer mesocosm experiment so that community-level responses could be compared. Four 20-day seasonal mesocosm experiments were undertaken between the winter of 1996 and autumn of 1997 to investigate seasonal variations in planktonic community responses to PCP. The mesocosms (860 L) were deployed in Lake Rotomanuka and established using water and planktonic organisms from the lake. Single applications of technical grade PCP dissolved in 95% ethanol were applied to the microcosms and mesocosms in a regression design experiment employing seven unreplicated nominal PCP concentrations (4, 10, 24, 36, 54, 81 and 121 μg.L⁻¹), with solvent and non-solvent controls. Temperature, dissolved oxygen, pH, conductivity, chlorophyll a, zooplankton and phytoplankton abundance were monitored regularly during the microcosm and mesocosm experiments. PCP concentrations were measured over a 40-day period in the spring mesocosm experiment so that nominal concentrations could be confirmed and rates of decay calculated. The effects of treatment and time on plankton species composition in the microcosms and mesocosms were analysed using redundancy analysis (RDA), with Monte Carlo permutation tests to identify significant treatment effects on each sample day. Daily EC50 estimates were calculated using total copepod % change in abundance so that effect concentrations could be compared to results from the laboratory copepod tests. In the laboratory tests, C. lucasi was the most sensitive of the three copepods with 48 h LC50 values of 52 and 106 μg.L⁻¹ PCP for nauplii and adults respectively. B. delicata nauplii and M. leuckarti adults were least sensitive (227 and 173 μg.L⁻¹ respectively). Summer collected C. lucasi adults at 22°C were least sensitive (106 μg.L⁻¹), while winter and spring collected adults at 12 and 17°C were most sensitive (70 and 71 μg.L⁻¹ respectively). PCP delayed metamorphosis in all species. C. lucasi were most sensitive (ChV 14.1 μg.L⁻¹) and M. leuckarti least sensitive (ChV 104.3 μg.L⁻¹). Community-level responses to PCP were similar in the microcosms and summer mesocosms with treatment effects apparent at concentrations as low as 10 and 24 μg.L⁻¹ in both systems. There was a slight delay in response in the microcosms, with copepod EC50 values on days eight (85 μg.L⁻¹) and twelve (90 μg.L⁻¹) in the microcosms similar to those on days two (96 μg.L⁻¹) and four (89 μg.L⁻¹) in the mesocosms. Following observation of a solvent effect, as reduced oxygen concentrations in the microcosms and mesocosms, a microcosm study was undertaken to investigate the effect of solvent concentration (8, 16, 32, 63, 94, and 125 μL.L⁻¹ ethanol) on pH and dissolved oxygen. Declines in oxygen and pH occurred in all treatments, although a recovery towards pretreatment levels was observed in the two lowest concentrations after eleven days. Planktonic community sensitivity to PCP in the mesocosms varied seasonally. In the RDAs treatment effects were identified at concentrations of 24 μg.L⁻¹ in summer, 36 μg.L⁻¹ in spring and autumn, and 54 μg.L⁻¹ in winter. With the exception of Cryptomonas, which responded positively to PCP, all other taxa were most sensitive in winter and spring, with copepods showing the greatest sensitivity. The seasonal variation in mesocosm copepod 48 h EC50 values, which ranged from 47 to 90 μg.L⁻¹ in spring and summer respectively, corresponded with changes observed in the laboratory tests using seasonally collected C. lucasi. Measured PCP concentrations on day one in the spring mesocosms were similar to predicted nominal concentrations. The mean half-life of PCP was ~21 days. In general, copepod LC50 values from laboratory tests were comparable to PCP concentrations causing acute responses in the microcosms and summer mesocosms. Copepod chronic values from sublethal tests were also similar to microcosm and mesocosm EC50 values and concentrations causing community-level treatment effects.