Drought is among sweetpotato production constraints in sub-Saharan Africa. Two studies were conducted on 15 F 1 sweetpotato families (G1-G15) generated using a half-diallel mating scheme of six parents. The first experiment was conducted at Kenya Agricultural Research Institute (KARI), Kiboko, using split plot design under drought stress and no drought stress replicated twice and repeated thrice between January 2012 and June 2013. The second study was conducted in the screen house at KARI, Muguga using randomized complete block design. General combining ability (GCA) and specific combining ability (SCA) effects for storage root yield (FSR), total biomass (BIO), harvest index (HI), marketable number of storage roots (MNR) and root dry matter (%RDM) were significant (P B 0.05) under both conditions. GCA/SCA ratio for FSR, HI, and %RDM under both conditions, ranged 0.51-0.76, thus, additive gene effects were more important than non-additive genes effects. G15, G5 and G7 had the highest significant (P B 0.05) FSR SCA effects under drought stress while G15, G7, and G12 had the highest under no drought stress. Progenies G8-8, G15-5 and G15-8, had the highest FSR mid and best parent heterosis ranging 117.8-269.6 % under drought. Drought susceptible parents P3, P4 and P5 had the highest yielding crosses under drought,which were also high yielding under no drought [G15 (P3 9 P5) and G5 (P4 9 P5)]. Thus, these parents probably were carriers of the drought tolerant genes (heterozygous recessive). This suggests that the drought tolerance alleles could be homozygous recessive, which may be confirmed in further studies.