Inbreeding effects in two pearl millet [Pennisetumtyphoides (Barm) Stapf] populations were estimated. From 208 groups of full-sibs, developed from each of two randommated populations, selections were made on the basis of best -8 and poor -8 yield performance and were designated as beat and poor subpopula-tions. Four inbred subpopulations, representing 0–2749, 0- 1492, 0–0854 and 0–0511 levels of inbreeding in each population were obtained by using full-sib family perse, intermating of 1, 2, 4 and 8 full-sib families, respectively. There was a gradual decline in mean performance with increased homozygosity from F=0.0511 to 0–2749 for all the characters except for days to flower. The differences among inbreeding levels between the subpopulations of two composites were significant. The inbreeding depression was maximum with 0–2749 coefficient of inbreeding. The results suggested that the use of a single full-sib family in forming a new population resulted in greater inbreeding depression and, therefore, two or more families should be used in intermating.
