Time of sowing has effects on growth and yield of most crops. On‐station experiments were conducted at two locations in Benue State, Nigeria, with the aim of studying the effect of sowing date on grain yield of rice. Five upland rice varieties were sown in a randomized complete block design with three replications on 15 June, 30 June, and 15 July in 2002 and 2003. Result showed significant difference in yields for sowing dates, but interaction effect of sowing dates and variety were statistically insignificant. The sowings on 30 June resulted in the highest yield of 3.32 Mg/ha in 2003 at Otobi, while the sowings on 15 July gave the least yield of 1.46 Mg/ha in 2002 at Makurdi. The regression of sowing date and yield showed a declining yield trend with delay in sowing dates for both locations. The Makurdi location had its predicted maximum yield with sowing on 15 June with yield of 2.70 Mg/ha. At Otobi the highest yield of 3.30 Mg/ha was on 23 June. The comparison between estimated and observed yield for both years for Otobi and Makurdi were not significant. A 12‐day delay in sowing beyond optimum sowing date resulted in a 5% decrease in predicted yield at Makurdi while at Otobi, a delay of 15 days resulted in a 5% decline in yield. These results enforce the need to rice farmers in Benue State to sow in June.