•We established a novel methodology for risk assessment of resistance development.•Risk assessment employed selection effect, resistance prevalence and stability.•Effect of antibiotic load on resistance risk was seasonal- and category-dependent.•Source tracking by fate and transport modeling revealed antibiotic load hotspots.•We proposed resistance-risk-targeted load reduction strategy in lake-river complex. Antibiotic stewardship is hindered by a lack of consideration for complicated environmental fate of antibiotics and their role in resistance development, while the current methodology of eco-toxicological risk assessment has not been fully protective against their potential to select for antibiotic resistance. To address this problem, we established a novel methodologic framework to perform comprehensive environmental risk assessment of antibiotics in terms of resistance development, which was based on selection effect, phenotype resistance level, heteroresistance frequency, as well as prevalence and stability of antibiotic resistance genes. We tracked the contribution of antibiotic load reduction to the mitigation of environmental risk of resistance development by fate and transport modeling. The method was instantiated in a lake-river network-basin complex system, taking the Taihu Basin as a case study. Overall, antibiotic load posed no eco-toxicological risk but an average medium-level environmental risk for resistance development in Taihu Lake. The effect of antibiotic load on resistance risk was both seasonal-dependent and category-dependent, while quinolones posed the greatest environmental risk for resistance development. Mass-flow analysis indicated that temporal-spatial variation in hydrological regime and antibiotic fate together exerted a significant effect on antibiotic load in the system. By apportioning antibiotic load to riverine influx, we identified the hotspots for load reduction and predicted the beneficial response of resistance risk under load-reduction scenarios. Our study proposed a risk-oriented strategy of basin-scaled antibiotic load reduction for environmental risk control of resistance development. Display omitted