The interactive effects of contaminants and ultraviolet light (UV)-exposure on the incidence and types of abnormalities observed were measured in newly metamorphosed cane toads (Rhinella marina) from four Bermuda ponds contaminated with petrochemicals and metals. Abnormalities were compared in toadlets that were field-collected, reared in predator exclusion cages, reared in laboratory microcosms exposed to control media or corresponding pond media, and reared in laboratory microcosms exposed to UV-light and control media or media from two ponds. Percent abnormal for field-collected, cage-reared, and microcosm-reared toadlets were equivalent per site and ranged between 14% and 63%. All treatments produced similar limb abnormalities but the percentage of hind versus forelimb defects was statistically greater only in field-collected toadlets. UV-exposed control media did not induce abnormalities in larvae exhibiting no maternal effect, and did not alter the types of abnormalities observed in larvae exhibiting a maternal or latent effect. Site media treatments without UV exposure induced significant cephalic and limb abnormalities, proved additive to the observed maternal/latent effect, and produced limb defects predominantly in forelimbs. Concurrent exposure to site media and UV-light induced similar types of abnormalities but a significantly higher percentage of hind limb abnormalities (68-89%) than exposure to site media alone (7-13%). Our results suggest that the types of abnormalities expressed were principally determined by direct and/or transgenerational contaminant exposure, but that UV-light exposure caused limb abnormalities to occur primarily in the hind limbs, mirroring field observations. Our field observations also suggest that ectromelia and brachydactyly in some field-collected specimens may be predator-induced.