In this study, the acute copper sensitivity of field-collected cladoceran species was determined using their natural surface waters and a standard reconstituted test water as test medium. A total of 43 species were collected on two occasions from six different sites, representing different water types and chemistries in Europe. The collected species belonged to four different families (Daphniidae, Bosminidae, Macrothricidae, Chydoridae) and 11 different genera (Daphnia, Ctenodaphnia, Ceriodaphnia, Simocephalus, Scapholeberis, Alona, Acroperus, Chydorus, Eurycercus, Disparalona, Pleuroxus). In acute experiments with immobilization as end point, the 48-h median effective concentrations (48-h EC50) for the cladoceran species ranged from 5.30 to 70.6 microg of Cu L(-1) in standard test water and from 9.60 to 853 microg of Cu L(-1) in natural waters. The mean site sensitivity (the geometric mean of 48-h EC50 values of species within a community) ranged from 10.1 to 27.4 microg of Cu L(-1) in standard water and from 16.4 to 281 microg of Cu L(-1) in natural water. This indicates that bioavailability is more importantthan inter-community (species composition) differences in determining the variability of copper toxicity across different aquatic systems. For the four surface waters that had a pH within the range for which the acute Daphnia magna biotic ligand model (BLM) has previously been successfully validated, the BLM predicted 48-h EC50 values for 27 of the 28 tested cladoceran species within factor of 2 of the observed values. For the same sites, all community sensitivities were predicted within a factor of 2.3. The BLM was clearly over-protective for the two acidic surface waters tested. Hence, the BLM can be considered a valuable tool for estimating the potentially harmful effects of copperto natural cladoceran communities, but more research will be needed for acidic surface waters.