Research has shown that hybrids tend to yield higher than inbred lines and hybrid-inbred line mixtures. Variable levels of heterosis are reported in literature. In Kenya yield advantage of up to 54% has been reported in extreme drought situations. Hybrids have therefore been suggested as an alternative to increasing grain yields in the eastern Africa region. No sorghum hybrids were developed in the past because sorghum farmers were viewed as too poor to afford hybrid seed. However, current social economic studies indicate that this may no longer be true. The present study aims to breed sorghum hybrids that will confer to farmers an economic yield advantage over the current open-pollinated varieties. The sorghum breeding program of the Kenya Agricultural Research Institute (KARI) has no male sterile sorghum lines on which to base hybrids. Therefore, male sterile lines for this study were acquired from collaborators in the International Sorghum and Millet Program (INTSORMIL) group of universities in the USA (from Kansas, Indiana, Texas) and the International Centre for Research in the Semi-Arid Tropics (ICRISAT)-Zimbabwe (selected after 20 years of regional testing). Local germplasm is expected to enhance hybrid vigour and adaptation of the hybrids with possible genetic distance. The introduced germplasm consisted of 94 parental lines, 41 pollen parents and 53 male sterile parents. They were screened together with 27 pollen parents from Kenya for co-adaptation at the University of KwaZulu Natal, South Africa (29o40'S, 30o25'E) and at Kiboko, Kenya (2o12'S, 37o43'E, 915 m altitude) on a Luvisol soil. From these tests 66 pollen parents and 51 male sterile lines were hybridized in a factorial mating design in Kenya during the January–May 2003 period. In a single-crossing block almost 2000 hybrid combinations could be realized, and 900 hybrids produced sufficient seed to plant four trials: high density (9 plants m–2) and low density (1 plant m–2) over two irrigation regimes (30 mm water weekly until 60 days after planting and up to 100 days after planting). Dry head weight from the high-density, high-irrigation regime showed significant variation in heterosis and specific combining ability in the hybrids. Male sterile parents exhibited significant variation in general combining ability. Yield advantage of the highest-yielding hybrid over the best KARI inbred line was 50%. From the seed count, it is concluded that the male sterile parent set is more variable than the male fertile set. It is inferred that high-yielding hybrids can be selected in a high-density, high-irrigation regime either by selecting high general combining ability (GCA) in parents or by selecting specific parent combinations that have high yield. These provisional results indicate that hybrids may be relatively easily produced in Kenya and potentially may increase sorghum yield up to 50% over the open-pollinated varieties. Economics of producing the best hybrids will tell whether or not there is a place for hybrids in Kenya.