Urban stormwater runoff is a globally significant threat to the ecological integrity of aquatic habitats. Green stormwater infrastructure methods such as bioretention are increasingly used to improve water quality by filtering chemical contaminants that may be harmful to fish and other species. Ubiquitous examples of toxics in runoff from highways and other impervious surfaces include polycyclic aromatic hydrocarbons (PAHs). Certain PAHs are known to cause functional and structural defects in developing fish hearts. Therefore, abnormal heart development in fish can be a sensitive measure of clean water technology effectiveness. Here we use the zebrafish experimental model to assess the effects of untreated runoff on the expression of genes that are classically responsive to contaminant exposures, as well as heart-related genes that may underpin the familiar cardiotoxicity phenotype. Further, we assess the effectiveness of soil bioretention for treating runoff, as measured by prevention of both visible cardiac toxicity and corresponding gene regulation. We find that contaminants in the dissolved phase of runoff (e.g., PAHs) are cardiotoxic and that soil bioretention protects against these harmful effects. Molecular markers were more sensitive than visible toxicity indicators, and several cardiac-related genes show promise as novel tools for evaluating the effectiveness of evolving stormwater mitigation strategies.