This study quantified the magnitude of heterosis in pearl millet (Pennisetu
m glaucum) topcross hybrids produced by crossing 16 diverse landraces and t
hree high yielding open-pollinating varieties on two homozygous male-steril
e lines. Hybrids and pollinators were grown in 12 year x location combinati
ons in India that were grouped into three zones. Genetic components of vari
ance quantifying the differences among these hybrids were estimated. The hy
brids showed a conspicuous heterosis for grain yield, earliness and biomass
yield but not for straw yield. The level and direction of heterosis for ti
me to flowering depended strongly on the earliness of the male-sterile line
. In the terminal drought stress zone hybrids made on the early maturing ma
le-sterile line 843A had the highest level of heterosis for grain yield (88
%). This was partly due to escape from terminal stress. In the other two zo
nes the heterosis for grain yield was on average 30%. Heterosis for biomass
yield and biomass yield per day was on average also positive in all three
zones. For all traits, except time to flowering and biomass yield per day,
pollinator effects were the only significant source of variation. Differenc
es between hybrids were mostly caused by additive genetic effects. Signific
ant amount of heterosis observed in landrace-based topcross hybrids for gra
in yield and other productivity-related traits suggested that substantial i
mprovement in pearl millet productivity in and environments can be obtained
by topcrossing locally adapted landraces on suitable male-sterile lines.