The influence of dissolved organic carbon (DOC) and water composition on the toxicity of copper to juvenile freshwater mussels (fatmucket, Lampsilis siliquoidea) were evaluated in natural and reconstituted waters. Acute 96‐h copper toxicity tests were conducted at four nominal DOC concentrations (0, 2.5, 5, and 10 mg/L as carbon [C]) in dilutions of natural waters and in American Society for Testing and Materials (ASTM) reconstituted hard water. Toxicity tests also were conducted in ASTM soft, moderately hard, hard, and very hard reconstituted waters (nominal hardness 45–300 mg/L as CaCO₃). Three natural surface waters (9.5–11 mg/L DOC) were diluted to obtain a series of DOC concentrations with diluted well water, and an extract of natural organic matter and commercial humic acid was mixed with ASTM hard water to prepare a series of DOC concentrations for toxicity testing. Median effective concentrations (EC50s) for dissolved copper varied >40‐fold (9.9 to >396 μg Cu/L) over all 21 treatments in various DOC waters. Within a particular type of DOC water, EC50s increased 5‐ to 12‐fold across DOC concentrations of 0.3 to up to 11 mg C/L. However, EC50s increased by only a factor of 1.4 (21–30 μg Cu/L) in the four ASTM waters with wide range of water hardness (52–300 mg CaCO₃/L). Predictions from the biotic ligand model (BLM) for copper explained nearly 90% of the variability in EC50s. Nearly 70% of BLM‐normalized EC50s for fatmucket tested in natural waters were below the final acute value used to derive the U.S. Environmental Protection Agency acute water quality criterion for copper, indicating that the criterion might not be protective of fatmucket and perhaps other mussel species.