Abstract In response to uncertainties and complexities in the irrigation water management systems, this study presents an improved interval-based fuzzy credibility-constrained programming (IIFCCP) model for optimizing the allocation of irrigation water resources under uncertainty. Mathematical techniques of interval parameter programming (IPP), improved fuzzy credibility-constrained programming (IFCCP) and interval crop water production functions (ICWPFs) are incorporated into a modeling framework for irrigation planning. Thus, this model has the following several advantages in irrigation water management. First, it can not only address interval and fuzzy uncertainties, but also analyze tradeoffs between irrigation water amount and agricultural production returns, to maximize the agricultural economic benefits. Second, it further improves upon existing studies on deterministic CWPFs by using the ICWPFs to reflect uncertain information in the practical applications. Third, it makes a novel attempt to explore weight coefficients (i.e. the proportion of possibility in estimating credibility) to represent various preferences of decision makers. Fourth, it provides more flexible strategies by considering various weight coefficients and credibility levels. Then, the proposed model is applied to a case study of irrigation water management in the middle reaches of the Heihe River Basin, northwest China. Therefore, optimal decision solutions for the allocation of irrigation water under different scenarios, i.e., the combinations of three water levels, four credibility levels and nine weight coefficients, will be obtained and analyzed. A greater system benefits will be achieved at a high weight coefficient and a low credibility level, especially weight coefficients are assigned as 0.8 and 0.9. Additionally, these solutions can provide detailed tradeoffs among system benefits, credibility levels and weight coefficients. Comparisons with the results from deterministic CWPFs and conventional FCCP method indicate that the IIFCCP method has an enhanced practical applicability in actual irrigation water management under uncertainty conditions in arid areas.