Cultivating hybrid rice variety has the potential to fulfill the ever-increasing demand for rice in Bangladesh due to its high yield. This study is intended to explore the determinants and impact of hybrid rice variety adoption on the yield and technical efficiency (TE) of Bangladeshi rural rice farms. The stochastic frontier production has been employed to approximate production frontier and TE using 15,952 rice plots from Bangladesh. This study has addressed self-selection into the hybrid variety choice by using propensity score matching (PSM) that adjusts observed selection bias, potentially influencing decisions to choose a hybrid variety and TE. The marginal effect of the Probit model reveals that farm size, education, subsidy card, irrigation access, irrigation machinery ownership, and non-farm business have positively while drought risk and household head age have negative influences on choosing a hybrid variety. We found that the hybrid rice variety adopters is significantly 6.24% more efficient and 12.03% higher frontier predicted yield than their counterparts. The average treatment effect on the treated also confirms that adopting a hybrid variety significantly impacts rice yield and TE. The primary and significant source of the varietal yield gap (19.3%) was differences in endowments (6.7%) and coefficients (12.1%). Fertilizer, irrigation, seed, clay, and clay loam soil have significantly contributed to explaining the difference and is critical in closing the existing yield gap. Policy implications include introducing hybrid rice varieties by motivating farmers through extension services, price policies, and irrigation services can help achieve desirable yield and enhance farm efficiency.