We conducted an ant-exclusion experiment in a citrus orchard to evaluate the overall impact of three ant species native in the Mediterranean, Pheidole pallidula (Nylander), Plagiolepis schmitzii Forel, and Lasius grandis (Forel), on populations of Aonidiella aurantii Maskell (California red scale). The ant-exclusion was carried out in four experimental plots from March 2007 to November 2008. Another subset of four plots, adjacent to the ant-excluded plots, was used as control. We measured scale densities and percent parasitism on fruits at harvest in 2007 and 2008. Additionally, we sampled the seasonal trend of the scale on twigs and fruits in both treatments during 2008. California red scale densities in the ant-excluded treatment began to be significantly lower than in the ant-allowed control in May (1 mo after ant activity began), and this difference increased until November. Thus, the effect of the ants on California red scale density seems to be accumulative. At harvest, scale densities on fruits were significantly lower in the ant-excluded treatment. However, percent parasitism on fruits was similar between treatments. Finally, scale densities on the fruits of the ant-allowed plots were positively correlated with the number of ants that climbed to the citrus canopy. These results suggest that increases of scale densities induced by Mediterranean ants depend on the intensity of the ant-activity on citrus canopies.