The available potassium in the rice soils representing six districts of Meghalaya was determined by five chemical extraction methods in order to find out the most reliable soil testing method for predicting response of rice crop to applied K for in the acid soils. Total K in the soils ranged from 1200-3950 mg kg-1 and exchangeable, non-exchangeable and lattice K fractions constituted 5.1, 8.6 and 85% of total K, respectively Distribution of various fractions of K in different soil groups followed in the order: Vertisols > Inceptisols > Ultisols > Alfisols > Entisols. Nitric acid extractable K recorded highest values of available K, ranging from 105 to 748 kg ha^-1, whereas the ammonium acetate extractable K varied from 71-403 kg ha^-1. Dry matter yield and K uptake widely varied from 13.45-55.77 g pot^-1 (30 x 30 cm) and 109-1008 mg pot-1 (30 x 30 cm) with a yield and uptake response of 20 to 110 and 27-247 %, respectively. Water soluble and exchangeable forms of K contributed 65, 68, 44, 66 and 44 percent variation to the extractable K by ammonium acetate, sulphuric acid, nitric acid, hydrochloric acid and calcium chloride, respectively. Water soluble and exchangeable forms of K showed highly significant relationship with K uptake of rice, indicated that these two forms significantly contributed to K nutrition of rice. Ammonium acetate extractable K was found superior to the other extraction methods as it showed highly significant relationship with plant growth parameters, with a critical limit of 194 kg ha^-1, below which the crop showed response to the applied K fertilizers. The H₂SO₄ extractable K was also found equally good as it showed highly significant relationship with plant growth parameters and various forms of K, with a critical limit of 210 kg ha-1.