Most of the agricultural researchers interested in statistical inferences that require multiple comparisons of means. The paper addresses the multiple comparisons procedures in crop variety trial, which is where multiple comparisons are a particular used in the Genstat statistical software. For multiple comparative assessment, grain yield (kg/ha) of mille genotypes was used. The results highlighted that Fisher's protected least significant difference (LSD) and Ryan-Einot-Gabriel Walsh were an appropriate procedure for multiple comparisons of means. Multiple comparison procedures in agricultural experiments are very useful for making the best decisions regarding to comparisons of treatment means in randomized complete block designs.