To control phytophagous mites on soybean crops in an economically viable way, it is necessary to quantify the occurrence of the mites on the leaflets. Estimating the number of mites cm(-2) on leaflets is more difficult because of their irregular distribution on the leaflet surface. Therefore, the aim of this study was to determine the count-area/leaflet and the number of soybean leaflets to quantify the mites. One hundred infested plants were randomly collected. One leaflet was removed from each plant and divided into 32 sections (1.0 cm(2) per section), arranged in four columns and eight rows, to count the mites (adults, nymphs and eggs). The ideal count-area size per leaflet (Xo) was estimated by the maximum curvature of the coefficient of variation method for each of the 100 leaflets. For a count-area of Xo size, we obtained the number of mites cm(-2) per leaflet and, using the bootstrap resampling method, we estimated the point and interval averages as well as the sample size for a pre-established error. We suggest that the determination of the evaluated area size on each soybean leaflet (20 cm(2) in this case) and the bootstrap resampling estimate of the appropriate number of leaflets (12 in this case) for a bootstrap confidence interval of four mites (adults + nymphs) cm(-2) is sufficient to standardize the sampling-procedures for quantifying mites on soybean leaflets.