Two components of seed (rice) production system viz.; seed harvest and storage were validated through onfarm trials with an objective to fine-tune existing and/or improved techniques based on location specific requirements. Panicle harvested (PH) seed was compared with that of traditional bulk harvest (BH) using two improved upland rice varieties (Vandana and Anjali) during 2004-05. Despite average yield increase of 6.1% (Anjali) to 14.9% (Vandana) the PH method was not accepted by the farmers due to higher labour requirement of the operation. So, a less labour intensive method of ‘crop area selection’ was evaluated and demonstrated during 2006-07 using the variety Anjali. Improved seeds resulted in significant grain yield increase to the tune of 12.5% (2.36 t ha-1) to 20.3% (2.31 t ha^-1) in on-farm and research station trials, respectively. Storage systems used by the farmers (locally termed as; Hatwa/Kothi/Gaza/Tokri) were comparable to hermetically sealed storage systems for seed viability, germination and seedling vigor. Based on the results and farmers’ opinion, the ‘crop area selection’ method followed by traditional method of seed storage was recommended for adoption. However, the farmers were advised to repeat panicle harvest once in 5-6 years, to regain seed purity.