For two decades, fish tumour surveys have been used to monitor habitat quality in the Chesapeake Bay (USA) watershed. Tributaries with sediments contaminated with polynuclear aromatic hydrocarbons (PAHs), known to cause liver neoplasia, were frequently targeted. Here, we compare surveys in brown bullhead Ameiurus nebulosus conducted in 2009-2011 in the tidal Potomac River watershed (including the Anacostia River) with previous surveys. Using logistic regression, we identified length and sex as covariates for liver and skin tumours. We reported a statistically significant decrease in liver tumour probabilities for standardized 280 mm Anacostia bullheads between the 1996 and 2001 samplings (merged collections: female-77.5%, male-43.0%) and 2009-2011 (female-42.2%, male-13.6%). However, liver tumour prevalence in bullheads from the Anacostia, Potomac River (Washington, DC) and Piscataway Creek (17 km downriver) was significantly higher than that for Chesapeake Bay watershed reference locations. The causes of skin tumours in bullheads are uncertain, requiring further research. The similar liver tumour prevalence in these three locations suggests that the problem is regional rather than restricted to the Anacostia. To monitor habitat quality and the success of pollution control actions, we recommend conducting tumour surveys on a 5-year cycle coordinated with sediment chemistry analyses.