The application of Rainfallâ€“Potential evapotrans- piration (P-PE) model to yam (Dioscorea rotundata) production was carried out in an on-farm yam trial during the 2006-2007 and 2007-2008 cropping seasons in Abeokuta, South-western Nigeria. The experiment was a 3 x 2 factorial arrangement of three varieties of yam (Efuru, Ise-osi and Oniyere), two planting dates (early and late). The resulting 6 treatments were replicated three times with 14 mounds in each plot, in randomized complete block design. The general model for selecting the planting date of each yam cultivar in the two experimental years was 0.1PE less than P less than 0.5PE. This was partitioned for the purpose of attaining optimal planting date into early and late, respectively as Î£(P-0.1PE) â‰¤ 0 designated as T1 and Î£(P-0.5PE) â‰¤ 0 as T2. For the rest of the season defined by P> PE, the physiological parameters and hydrothermal agro-climatic indices measured during the different phenological stages of yam grown were analyzed with respect to the various treatments. The result showed that model T1 defined as Î£(P-0.1PE) â‰¤ 10mm appeared as the best model that significantly (P less than 0.05) influenced emergence rate, phenological growth and tuber yield. The implications of the study for appropriate schedule of farm operations vis-Ã -vis agronomic practices for yam cultivation have been noted.