Environmental stressors that severely impact some species more than others can alter ecosystems and threaten biodiversity. Genotoxic stress, such as solar UV-B irradiance, may induce levels of DNA damage at rates that exceed repair capacities in some species but remain below repair capacities in other species. Repair rates would seem to establish toxicity thresholds. We used inbred Xenopus laevis tadpoles in the laboratory to test the hypothesis that balances between rates of induction of cyclobutane pyrimidine dimers (CPDs; the major UV-B photoproduct in DNA) and rates of CPD removal (repair) can determine UV-B toxicity thresholds. As rates of chronic UV-B irradiance were progressively increased by decreased shielding of lamps, survival decreased sharply over a relatively narrow range of dose rates. Apparent toxicity thresholds were associated with large increases in steady-state CPD levels. Induction at twice the measured removal (repair) rate produced sustained high CPDs and 100% mortality. Induction at one-half the removal rate resulted in negligible CPD levels and low mortality. Increased intensity of visible radiation available to drive CPD photoreactivation, mimicking interspecies variation in DNA repair capacity, reduced steady-state CPD levels and increased survival at UV-B dose rates that were previously toxic, resulting in increased thresholds of apparent toxicity. We suggest that threshold effects due to DNA repair should generally be considered in assessments of effects of genotoxic agents on species-specific population decreases and human health risks.