From 1993 to 1995 data sets were collected from four citrus groves in Valencia, Spain, to determine the distribution patterns of eggs and nymphs of Aleurothrixus floccosus (Maskell), Dialeurodes citri (Ashmead), and Parabemisia myricae (Kuwana) on leaves, and to develop reliable sampling plans for estimating densities of immature whiteflies. A. floccosus showed higher aggregation than the other two species. The dispersion index b from the Taylor power law did not vary between different developing stages for A. floccosus and D. citri, reaching overall values of 1.70 and 1.53, respectively. In P. myricae, b was 1.60 for eggs and N1, and 1.46 for the remaining nymphs. The minimum number of leaves to estimate the population density with a coefficient of variation of 0.25 for densities above 10 immature whiteflies per leaf was 40 for D. citri and P. myricae, and 250 for A. floccosus. Binomial sampling programs for the three species were rejected for pest management purposes due to the high sample sizes required. The enumerative procedure of counting the number of insects per leaf appears to be the most suitable method for D. citri and P. myricae. For A. floccosus an index of occupation (from 0 to 10) linearly related to the proportion of the leaf undersurface occupied by this insect was found to be reliable and time-saving. Examining 150 leaves with this index achieves the desired relative variation level of 0.25 for most population densities usually found in commercial groves.