The interaction between water regimes and potassium levels was studied, using two rice cultivars, i.e., Sakha 104 and Giza 178, at the Experimental Farrn of Rice Research and Training Center (RRTC), Sakha, Kafr El-Sheikh, Egypt, during 2003 and 2004 rice seasons. Four water regimes, namely, continuous flooding, continuous saturation, Irrigation every six and nine days as well as four potassium levels, namely, 0, 57, 86 and 114 kg K20/ ha were used. The experiments were laid out in a split split-plot design, with four replications. The water regimes were applied in the main plots and the cultivars were placed in the sub-plots. Besides, the potassium levels occupied the sub-sub plots. The main results revealed that no significant differences were obtained between continuous saturation and continuous flooding on grain yield and most of its attributes. Prolonging irrigation Intervals up to nine days resulted in about 17 % reduction in grain yield. Giza 178 produced a higher dry matter, grain yield and water productivity than Sakha 104. The results, also, showed that increasing K levels up to 86 kg K20/ha Increased significantly plant height, dry matter and grain yield and its attributes under prolonging irrigation intervals In both seasons. The present study indicated a reduced water input and an increased water productivity of rice grown under just-saturated soil conditions, compared with continuous flooded rice. Additionally, continuous saturation gave a similar grain yield as that of continuous flooding with only 3-4% reduction in grain yield. This means that almost 20-25% of irrigation water could be saved, if all fanners followed this practice. As a result, a high water productivity could be obtained and about 1500 million LE. could be added to the national income.