Australian tropical freshwaters can experience extreme seasonal variability in rainfall and run off, particularly due to pulse events such as storms and cyclones. This study investigated how seasonal variability in dissolved organic matter (DOM) quality impacted the chronic toxicity of copper to a tropical green alga (Chlorella sp.) in the presence of two concentrations of DOM (low: ∼2 mg C/L; high: ∼10 mg C/L) collected from three tropical waters. Copper speciation and lability were explored using diffusive gradients in thin-films (DGT) and modelled maximum dynamic concentrations (cdynmax) using data derived from the Windermere Humic Aqueous Model (WHAM VII). Relationships between copper lability and copper toxicity were assessed as potential tools for predicting toxicity. Copper toxicity varied significantly with DOM concentration, source and season. Copper toxicity decreased with increasing concentrations of DOM, with 50% growth inhibition effect concentrations (EC50) increasing from 1.9 μg Cu/L in synthetic test waters with no added DOM (0.34 mg C/L) up to 63 μg Cu/L at DOM concentrations of 9.9 mg C/L. Copper toxicity varied by up to 2-fold between the three DOM sources and EC50 values were generally lower in the presence of wet season DOM compared to dry season DOM. Linear relationships between DGT-labile copper and dissolved copper were significantly different between DOM source, but not concentration or season. Modelled cdynmax consistently under-predicted labile copper in high DOM treatments compared to DGT measurements but performed better in low DOM treatments, indicating that this method is DOM-concentration dependent. Neither speciation method was a good surrogate for copper toxicity in the presence of different sources of natural DOM. Our findings show that DOM source and season, not just DOM concentration, affect copper toxicity to freshwater biota. Therefore, DOM quality should be considered as a toxicity-modifying factor for future derivation of bioavailability-based site-specific water quality guideline values.