Field trial were conducted in 1998 and 1999 growing seasons to determine the effect of time and frequency of weeding on the grain yield of an upland rice. The experiment involved five treatments, which were not weeding, weeding at  15DAS, weeding at 30 DAS, weeding at 15 and 45DAS and weeding at 15,30 and 45DAS. The effects of time and frequency of weeding indicated that treatments D that were weede at 15 and 45DAS had the highest grain yield with a mean of 2034kg/ha-1.This was followed by treatments E that were weeded at 15,30 and  45DAS with an average mean of 1882.85 g/ha-1. Based on the overall grain yield, treament D that were weeded at 15 and 45 DAS had an average yield of 7.45% over treatment E that were weeded at 15,30 and 45DAS,31.31% over treatments C that were weeded at 30 DAS, 59.75% over treatments B that were weeded at 15 DAS and 69.89% over the control treatments A that were not weeded at all.