The objective of this study was to develop an easily replicable sampling methodology using very high spatial resolution (VHSR) optical imagery to study the effect of landscape composition on crop pest incidence and biological control. The methodology was developed for the millet head miner (MHM), Heliocheilus albipunctella (de Joannis) (Lepidoptera: Noctuidae), a key pest of millet in Senegal (West Africa). The sampling plan was developed according to two main hypotheses: (i) pest incidence increases with millet abundance in the landscape, and (ii) biological control increases with the abundance of semi-natural habitats in the landscape. VHSR satellite imagery (<1¿m) provided from a Pléiades sensor was used to map and to quantify the landscape elements. Covering a square region of 20¿×¿20¿km, a hierarchical, broad-scale land cover map focusing on crop (millet and peanut crops) and tree (tree vegetation) categories was produced and validated with ground truth data. Then, the landscape variables (tree density index and millet crop density index) were calculated based on a regular grid of 100¿ha for each cell size covering the study area; the variables were then split into three density classes (low-medium-high) representative of the full landscape heterogeneity and combined into nine landscape patterns. Finally, according to sampling capacity, track accessibility, and statistical constraints, 45 field sites, including five replicates for each landscape pattern, were validated and selected for pest monitoring.