The contamination generated by industries in the municipality of Cubatão, in the coast of São Paulo state, has been causing damages to the structure of the surrounding Atlantic forest along the last decades. The Atlantic forest houses a high amphibian diversity and these animals may be good environmental indicators. The aim of this study was to detect possible effects of this contamination in frogs. Our hypothesis is that there is a gradient of contamination as the distance of the populations studied varied in relation to the source of pollution. Populations of Cycloramphus dubius (habitat specialist) and Rhinella ornate (habitat generalist) were sampled, in the regions, Mogi, Quilombo, and Itatinga, located at increasing distances from Cubatão. Each individual captured was submitted to measurements such as body mass and length, further used to calculate the body index (estimate of specimen body condition), besides measurements of the relative mass of organs and parasite load in lungs and intestine. These variables were compared among populations from different sampled areas. Specimen tissues and soil samples were tested for contaminant presence. At Mogi, the closest location from Cubatão, individuals of C. dubius and R. ornate presented larger liver and fat body mass for the same body index. In C. dubius, the Quilombo population, the intermediate location for pollution, presented smaller liver and fat body mass and larger loads of lung parasites. For R. ornate, the Mogi population presented larger liver, kidney and fat body mass, as well as larger spleen mass for the same parasite load. The locality where tissue and soil contamination was greater changed with contaminant type. This work presented conflicting results so that the hypothesized relationship between putative effects of contamination and distance from industries was not clear. Perhaps the constant water flow in the habitat of C. dubius does not allow pollutants to accumulate, and the differences found at Quilombo could be related to others factors of the species ecology. For R. ornate, perhaps differences found at Mogi reflect contamination from local sources. Finally, perhaps the contaminant analyses performed were not sufficient to demonstrate contamination differences among localities. Differently from expected, the generalist species seems to be more affected by contamination, showing the importance of considering species ecology in this kind of study.