ABSTRACT
Three resistant sorghum cultivars used as males were crossed to three susceptible sorghum cultivars used as females in a factorial mating system. The resultant nine F, hybrids were advanced to F2 generation and screened for resistance by artificially infesting each sorghum panicle with 10 pairs of head bugs (Eurystylus oldi (Poppius)) using the 'no choice' head cage method. The best 10% of F2 genotypes were advanced to F3 and F4 generations. The six parents, nine F,s, nine F2s, 25 F3 lines and 25 F4 lines were evaluated for resistance to E. oldi. Randomized complete block design (RCBD) was used, while North Carolina design II was the breeding design used. The analysis of variance revealed highly significant effects (P=0.01) among entries (nonsegregating and segregating populations), crosses, F2 generations, F3 lines and F4 lines for all the traits measured that is; number of head bugs, grain damage rating, floaters percentage, germination percentage, 1000-grain weight and grain yield. The parents also showed highly significant effect (P=0.01) for all the traits except grain yield. Estimates of genetic component of variance O2 A, O2 D and O2A/D ratio revealed the preponderance of non-additive gene action for all the traits measured. High narrow-sense heritability estimates were recorded for all traits in F4 lines, F3 lines, F2 generations and the F1S, except for number of head bugs (47.00%) in the F1. High narrow-sense heritability estimates were also recorded at F2 generations, F3 and F4 lines for all the traits measured.