In Cameroon, cocoa (Theobroma cacao L.) is traditionally grown in thick and multi-strata agroforestry systems. In such highly diversified ecosystems, cocoa production is seriously impacted by insect pests, especially by the cocoa mirid Sahlbergella singularis. Shade regulation is an ordinary management strategy for cocoa mirids, yet very few studies were able to assess the real effect of shade on mirid populations in cocoa agroforests. The main objective of this study was the validation of shade management strategies, by measuring the impact of shade on the spatial distribution of populations of S. singularis. For two consecutive years, mirid densities were assessed by a knock-down sampling method, in three 2025 m2 plots, located in three different agroecological zones in the Centre region of Cameroon. Plot shade was characterized by the stand density, the total basal area and the crown cover of shade trees. Moreover, light conditions were assessed using hemispherical photography. Geostatistical and mapping procedures were used to display shade layout and spatial distribution of mirid populations. Populations of S. singularis were highly aggregated in the plots. Semivariance analysis and kriging visualised the spatial dependence of mirid densities. Clearly distinguishable mirid pockets of 20 to 30 adjacent infested cocoa trees were identified in two of the three plots. The high diversity of shade tree species and the large variability in density and size of shade trees resulted in a considerable heterogeneity of plot light conditions. Percentage transmitted light varied from 9.4 % to 80.1 % in the most heterogeneous plot. For two of the three plots, mirid pockets were aggregated in those areas where light transmission was highest. In the third plot, relatively high mirid densities and the presence of alternative host plant resulted in a more homogeneous distribution. For the first time, we established a quantifiable link between light intensity at cocoa canopy level