The special statistical technique developed by Fisher for examining the influence of rainfall on wheat yield at Rothamsted has been employed to study the effect of rainfall distribution on the yield of wheat crop at the different crop-weather observatory stations in India. The study clearly establishes the influence of rainfall distribution on yield. About 75 % of the total variations in yield has been accounted for by the rainfall distribution. The response curves revealed that additional rainfall above the average during a period of about a month prior to sowing and during the deriod of germination is, in general, beneficial to the crop, whereas the same during the tillering phase is detrimental. IN TRODUCTION FISHER (1924) developed a special statistical technique for examining the effect of rainfall on the annual yield of wheat at Rothamsted. With the method evolved he estimated linear regression of wheat yields on rainfall distribution throughout the year and found the response curves which give the expected change in yield for an additional inch of rainfall above the average falling at any time of the year. The same method of analysis was applied by KALAMKAR and SATAKOPAN (1940) to examine the influence of rainfall on cotton yields at Government Experimental Farms at Akola and Jalgaon. NAm and BOSE (1945) generalised the above method to examine the influence of humidity and temperature on the yield of cotton at Sarkand (Sind). ACHARYA et al. (1960) used the method to examine the influence of rainfall on the sugar-cane yield at Pusa (Bihar). The same technique has been adopted in the present study to examine the influence of rainfall distribution on the wheat yield at cropweather observatories at Parbhani (latitude 19Β°16'N, longitude 76Β°47'E), Niphad (latitude 20Β°06'N, longitude 74Β°07'E), Powerkhera (latitude 22Β°44'N, longitude 77Β°42'E), Dharwar (latitude 15Β°27'N, longitude 76Β°06'E), Jalgaon (latitude 21Β°03'N, longitude 75Β°34'E), where data for approximately 15 years are available.