The potential of pollen selection as part of the breeding efforts to increase chilling tolerance in chickpea was investigated. This alternative approach to apply selection pressure at the gametophytic stage in the life cycle has been proposed widely, but there are no reports of the technique being implemented in a crop improvement program. In this paper, we describe how we developed a practical pollen selection technique useful for chickpea improvement.Pollen selection improved chilling tolerance in crossbreds compared with the parental chickpea genotypes and compared with progeny derived without pollen selection. This is backed up by controlled environment assessments in growth rooms and by field studies. We also clearly demonstrate that chilling tolerant pollen 'wins the race' to fertilise the ovule at low temperature, using flower color as a morphological marker. Overall, pollen selection results in a lower threshold temperature for podding, which leads to pod setting two to four weeks earlier in the short season Mediterranean-type environments of Western Australia. Field testing at multiple sites across Australia, as part of the national crop variety testing program, shows that these breeding lines have a significant advantage in cool dryland environments.The major factors which affected the success of pollen selection are discussed in the paper, from generation of variability in the pollen to a means to recover hybrids and integration of our basic research with an applied breeding program. We conclude that chilling tolerance observed in the field over successive generations are the result of pollen selection during early generations.