The over-consumption of foods with rapidly digestible carbohydrates is one of the factors for the onset of diabetes. Pearl millet is amongst cereals which appear to be beneficial against diabetes, due to its slower starch digestion properties. Starch is a carbohydrate which is digested in two phases corresponding to rapidly (RDS) and slowly (SDS) digestible starch. RDS is responsible for a rapid glucose release leading to increased glycaemic index within 20 minutes after ingestion, whereas SDS allows for a slower release of glucose over two hours. Resistant starch (RS), which is not digested, has health benefits including lengthening the period of satiety. Due to rapid rise in numbers of diabetics around the globe, it is getting important to analyse starch digestibility of food grains. However, the classical biochemistry protocols for starch digestibility analysis by Englyst (1992), which are still followed widely, are time-consuming. Here we report a modified starch digestibility method with improved throughput that is compatible with large-scale analyses of entries, such as Genome Wide Association Studies. We simplified and miniaturized the assays from 800 mg analysed in 50 mL tubes down to 100 mg in 5 mL tubes. This allowed doubling the number of samples passed in one day (36 samples), reducing the number of operators from two to one, and reducing the number of experimental hours, thus allowing for computer analysis to be done in the same day. An example of the type of results obtained on pearl millet seed germplasm using the original and the modified protocols will be shown. With two replicates, we obtain sufficient reproducibility so as to select for genotypes with lower RDS, higher SDS and higher RS.