Estimates of forest leaf litter frog density, mass, richness and diversity given by the widely used 8 m × 8 m large plot method (LPM) were compared with estimates obtained by a newly proposed method (small 2 m × 1 m plots with leaf removal; SPLR). The study site was an undisturbed area of the Atlantic Rainforest of Ilha Grande, an island located in the south of Rio de Janeiro State, Brazil. Twenty-four LPM (totalling 1536 m2 of forest floor) and 90 SPLR (totalling 180 m2 of forest floor) were performed. The estimates obtained by the two methods differed markedly, indicating that even using a much smaller sampling area (11.7% of that of LPM), SPLR gave frog density estimates six times higher, and frog mass estimates approximately 2.5 times higher than estimates provided by LPM. The species richness and diversity obtained by the two methods were similar, despite the fact that the total area sampled with SPLR was much smaller. These data suggest that LPM may underestimate the abundance and biomass of leaf litter frogs in a given area.