|dc.description.abstract||The genetic control of resistance to the cassava mosaic disease (CMD) in some African landraces, their relationship with the widely used resistant genetic stock, clone 58308, and molecular markers associated with resistance to CMD in the landraces were evaluated in this study. The F-i progenies of 54 cassava crosses and their parents (clone 58308, six improved clones and 15 African landraces) were evaluated in two genetic experiments in three environments in Nigeria. Genetic analysis revealed that additive gene effect was more important in the genetics of resistance to CMD among the landraces, while both additive and non-additive gene effecsts were important for clone 58308. The segregating Fi crosses exhibited varying levels of resistance and susceptibility to CMD suggesting polygenic inheritance. Resistant phenotypes were detected in the crosses involving susceptible parents which suggests that resistance to CMD in the clones studied is due to recessive genes and susceptible phenotypes in crosses involving resistant parents suggested that the resistance genes are non allelic. Positive transgressive segregants were also detected in some crosses. The number of effective factors for resistance to CMD ranged from two to seven and was contributed by both parents in a cross. Significant differences in the mean distribution of F-i progeny disease severity scores further revealed allelic differences between three improved clones and some landraces, and that expression of resistance is influenced by the nature of the female parent. Bulk segregant analysis (BSA) showed that an SSR marker SSR30-180 was associated with CMD resistance in a cross between the susceptible clone TMSI30555 and the resistant landrace TME7. Linkage analysis revealed that SSR30-180, which was donated by TME7, was 14.9 cM from a putative CMD resistance locus, CMDRL. Marker-trait association detected by regression analysis further showed that markers, SSR30-180, E-ACC/M-CTC-225, SSR119-A and SSR119 accounted for 57.41%, 22.5%, 32.24% and 37.83% of the total phenotypic variation respectively, for resistance to CMD in the mapping population. The results further showed differences between TME7 and six out of the 15 resistant clones used as parents and checks in the genetic experiments with respect to SSR30-180, suggesting that different alleles are involved in resistance to CMD. Two susceptible clones TME31 and TME117 however, had the marker, which suggests that SSR30- 180 is not tightly linked to resistance, thus there is a need to saturate the map with more markers. These findings are useful in the selection of cassava parental clones for resistance breeding.