The photothermal response of three Kabuli chickpea (Cicer arictinum L.) cultivars, at different Growth stages. to eight irrigation treatments in 1998/99 and four irrigation treatments in 1999/2000 was studied on a Wakanui silt loam soil in Canterbury. New Zealand (43degrees38S, 172degrees30E). The rate of development from emergence to flowering (e-f) and sowing to harvest maturity were strongly and positively associated (R-2=0.87, P<0.001) with mean temperature during those periods. All phenological stages considered (sowing to emergence, e-f, flowering to podding, podding to physiological maturity and physiological maturity to harvest maturity) depended upon accumulated thermal time (T-t) above a base temperature (T-h) of 1 degrees C. An accurate prediction of time of flowering was made based on an accumulated mean T requirement of 629 degrees Cdays from e-f (R-2=0.91, P<0.001). Fully irrigated crops had higher maximum dry matter accumulation (maxDM; 1093 g/m(2)), duration of exponential growth (DUR; 99 days), weighted mean absolute growth rate (WMAGR; 12.2 g/m(2) per day) and maximum crop growth rate (MGR: 17.1 g/m(2) per day). In 1998/99 the positive response of maxDM and MGR depended on a significant (P<0.01) interaction between irrigation and sowing date. The maxDM during the season was highly correlated with DUR and MGR (R-2=0.79 and 0.65). It is concluded that to maximize chickpea biological yield in the dry season of the cool-temperate subhumid climate of Canterbury, irrigation should extend across all phenological stages.