Low concentrations of elements in food can be measured with various techniques, mostly in small samples (mg). These techniques provide only reliable data when the element is distributed homogeneously in the material to be analysed either naturally or after a homogenisation procedure. When this is not the case or homogenisation fails, a technique should be applied that is able to measure in samples up to grams and even kilograms and regardless of the distribution of the element. An adaptation of neutron activation analysis (NAA), called large-sample NAA, has been developed and proven accurate and may be an attractive alternative in food research and mass balance studies. Like standard NAA, large-sample NAA can be used to measure both toxic and trace elements relevant for nutrition.