Weed competition remains a major constraint to productivity in rice (Oryza sativa L.)-based cropping systems, yet the ecological factors structuring weed communities across environments and management regimes remain insufficiently resolved in South Asia. This study examined weed community composition, diversity, and dominance in rice systems of Bangladesh and Nepal to assess how environmental conditions, crop rotation, and herbicide use shape weed assemblages. Weed community surveys were conducted in major rice-growing areas during 2016 and 2019 in Bangladesh and during 2019 in Nepal, covering 120 farmers’ fields per season across contrasting crop rotation systems (rice–fallow–rice and wheat (Triticum aestivum L.)–jute (Corchorus olitorius L.)–rice in Bangladesh; rice–wheat–fallow and rice–lentil (Lens culinaris Medik.)–fallow in Nepal) and management regimes with and without preemergence herbicide use. We studied species richness, diversity, relative abundance, and frequency and their association with cropping system, weed management, land type, soil texture, and water management. Weed diversity and community composition varied significantly by site, year, cropping system, and herbicide use. Fields managed without herbicides consistently exhibited higher species richness and evenness, while herbicide use was associated with reduced diversity and dominance by fewer species. In Bangladesh, the wheat–jute–rice system supported higher weed diversity (0.96 to 1.01) than the rice–fallow–rice system (0.84 to 0.91), whereas in Nepal, weed community structure differed more strongly among locations than among crop rotations. Dominant species in Bangladesh included yellow nutsedge (Cyperus rotundus L.), bermudagrass [Cynodon dactylon (L.) Pers.], barnyardgrass [Echinochloa crus-galli (L.) P. Beauv.], pickerelweed [Monochoria vaginalis (Burm. f.) C. Presl. ex Kunth], cosmopolitan bulrush [Bolboschoenus maritimus (L.) Palla]; syn.: Scirpus maritimus L.], and Paspalum spp., with relative abundance varying by land type and water regime. In Nepal, weed communities were characterized by recurring dominance of fimbry (Fimbristylis littoralis Gaudich.), ricefield flatsedge (Cyperus iria L.), C. dactylon, junglerice [Echinochloa colona (L.) Link], pimpernels (Lindernia spp.), and joyweeds (Alternanthera spp.), with shifts in abundance linked primarily to site-specific hydrological and soil conditions. Farmer identification of troublesome weeds closely matched measured dominance patterns. These findings provide an ecological foundation for refining site-specific weed management strategies.