Cropping systems are highly diverse, intensive, and complex in Bangladesh. Rice, maize, and wheat are the three major cereal crops grown in most of the farmlands with rice-rice, rice-wheat, and rice-maize being the major cropping systems. In recent years, the area under rice-wheat cropping system is declining while that under rice-maize is increasing. Continued increased production of rice is required to meet food demand due to growth in population and incomes, while increased production of maize is required for both human food and livestock feed as well as for other industries. The objective of this work was to quantify yield potential, yield gaps, and potential production for irrigated and rainfed rice and maize grown in Bangladesh. Target areas for estimation of yield potential and gaps (hereafter called reference weather stations; RWS) were selected based on crop-specific harvested area, climate zones, and location of weather stations with 20+ years of data. A total of 5 RWS were selected for rice and maize.  Yp and water-limited yield potential (Yw) were estimated using crop simulation models (ORYZA2000 and Hybrid Maize, respectively) based on RWS-specific actual weather, soil, and typical management practices (planting date, plant density and cultivar maturity) for the dominant annual crop sequences at each location (for example, rice-maize and rice-potato-maize). For each RWS, average yield potential was computed as a weighted average of the simulated yields for each crop sequence (based on their relative contribution to the total cropland area around the RWS) and the yield gap was estimated as the difference between RWS simulated Yp or Yw and actual (last 5-year) average on-farm yield reported for the district where the EWS was located. Results indicate that yield gaps in Bangladesh are highly variable across locations: for example, yield gap of irrigated maize ranged from 26 to 58% of the simulated yield potential, with  average national actual maize yield (5.8 Mg/ha) representing 57% of average simulated yield potential for irrigated maize in Bangladesh (10.1 Mg/ha). Results presented here highlight the importance of performing crop simulations within the context of the actual cropping system in order to obtain reliable and robust estimates of yield gaps, especially, in regions where 2+ cycles of the same crop are grown annualy in the same piece of land or where the same crop is grown in different annual crop sequences that co-exist within the same geographic area.