Abstract: Six field experiments of response of garlic yield to K application combined with N were conducted from 1998 to 2001 in Yellow-drab soil Jianghuai hilly land. The results showed that there was a remarkable response of garlic yield to potassium application, the increase of garlic bolt and clove yield was 28.7%~39.7% and 19.5%~28.2% with the average of 36.4% and 24.1% respectively compared with the control. The yield increase of K fertilizer was quite high at two rates of nitrogen application, and it was closely relative to the content of available K in soil and potassium application rate. The yield increase of N fertilizer was affected by the content of soil available N as well as potassium application and the content of available K in soil. The liner regression equations were developed with the method of stepwise regression analysis between the variables of the yield increase per K or N fertilizer of garlic bolt or clove (y) with that of the content of available K (x1) or N (x2) in soils and potassium or nitrogen application (x3) and (x4) respectively.