Abstract Antifouling paints on small to medium recreational vessels were first recognized as an important source of pollution in the 1970s. One of the principle biocides in these paints is Cu. Results from a field program in the Gold Coast Broadwater, Queensland, demonstrate a clear correlation between recreational boat numbers at anchorage sites and water column Cu concentrations, for both time-integrated 24 h DGT measurements ( n = 14, r = 0.815, p < 0.001) and measurements on composite 0.45 μm-filtered grab samples ( n = 14, r = 0.698, p < 0.01) collected every 4 h over 24 h. At boat numbers above 30 the 0.45 μm-filterable Cu concentrations were mostly above the relevant guideline value (1.3 μg l −1) but the DGT-reactive Cu concentrations were well below this value at all boat numbers studied. For three-day DGT deployments in Moreton Bay, Queensland, correlations between Cu levels and vessel numbers were not observed, possibly because of uncertain estimates of boat numbers. However, using a multi-factorial ANOVA, DGT-reactive Cu concentrations showed a significant effect for ‘sites with vessels vs. sites without vessels’ ( α = 0.10, p = 0.077) and for ‘sampling period’ (non-holiday weekdays, weekends, holiday weekdays, holiday weekends) at α = 0.10, p = 0.02. Cu levels in sediments at the same sites were strongly influenced by fraction of clay (<63 μm) material but also by whether the samples were collected at an anchorage or control site. Results from this study further support the view that Cu emissions from antifouling paints may become an important source in waters with high boat numbers and should be taken into account when designing management instruments for coastal waterways.