We present a method using high-angle annular dark field scanning transmission electron microscopy (HAADF-STEM) to determine the chemical composition of bi-metallic nanoparticles. This method, which can be applied in a semi-automated way, allows large scale analysis with a statistical number of particles (several hundreds) in a short time. Once a calibration curve has been obtained, e.g., using energy-dispersive X-ray spectroscopy (EDX) measurements on a few particles, the HAADF integrated intensity of each particle can indeed be directly related to its chemical composition. After a theoretical description, this approach is applied to the case of iron–palladium nanoparticles (expected to be nearly stoichiometric) with a mean size of 8.3 nm. It will be shown that an accurate chemical composition histogram is obtained, i.e., the Fe content has been determined to be 49.0 at.% with a dispersion of 10.4 %. HAADF-STEM analysis represents a powerful alternative to fastidious single particle EDX measurements, for the compositional dispersion in alloy nanoparticles.