A field experiment was conducted for three years during rainy (kharif) season of 2006–2008 at Almora, Uttarakhand to study the effect of different weed control methods on weed infestation, energy use efficiency, productivity and profitability of direct seeded rice (Oryza sativa L.) under rainfed conditions. On an average, a weed population of 627/m2 was recorded in the weedy check plots. The grain yield decreased by 78.5–94.8% due to season-long weed-crop competition as compared with hand weeding (20 and 40 DAS). The highest weed-control efficiency and loweet nutrient uptake by weeds was recorded with butachlor + hand weeding (40 DAS) closely followed by hand weeding twice (20 and 40 DAS). Both these treatments recorded higher grain yield (2.21–2.36 t/ha) compared to other treatments. Butachlor + weeding with wheel-hoe (40 DAS) was the next best treatment. This treatment required less energy input compared with all other weed control treatments. Butachlor + hand weeding and butachlor + wheel- hoe registered the highest energy use efficiency (10.23 and 9.75, respectively) and net returns ( 11,868 and 11,299/ha, respectively) compared with other treatments. The benefit: cost ratio was the highest (1.90) with butachlor + weeding with wheel-hoe closely followed by butachlor + hand weeding and butachlor + long-handled khurpi.