Five different methods of diallel analysis have been compared using data from a half-diallel cross of a fixed set of nine homozygous varieties and one set of their single cross progenies in chickpea. The interrelationships among various parameters obtained from these analyses are reviewed and the advantages and disadvantages of each method discussed. The analysis proposed by Gardner and Eberhart (1966) appears to be superior as in addition to gca and sca effects and variances it provides information on the additive effects of varieties and their average and individual contribution to heterosis in crosses.