Derived from an agricultural system model, a new cost-effective and water-saving decision support system for irrigation scheduling (DSSIS) was developed to control field irrigation of cotton (Gossypium hirsutum L.) at the Qira oasis, Xinjiang, China. DSSIS used the water stress factor and soil water content simulated by the Root Zone Water Quality Model (RZWQM2) to trigger irrigation and to calculate the irrigation amount. In this study, its potential effects on cotton photosynthesis and development were evaluated in field experiments (2017, 2018). Arranged in a randomized complete design with four replicates, the experiment contained three irrigation control methods [DSSIS–based (DSS), soil moisture sensor–based (SMS), and conventional experience–based (CE)] and two irrigation rates [full and deficit irrigation (75% of full irrigation)]. Based on the two–year results, the DSS significantly improved both leaf photosynthesis and root biomass. Compared to the SMS, the DSS and CE significantly enhanced net leaf photosynthetic rate (Pn) by 20.1% and 19.5%, respectively, stomatal conductance (Gs) by 38.9% and 36.5%, respectively, and transpiration rate (Tr) by 19.5% and 17.1%, respectively. The DSS and CE outperformed the SMS by producing higher aboveground and root biomass, plant height, leaf area, and weight and number of marketable bolls per plant (p ≤ 0.05). Compared to full irrigation, deficit irrigation caused 11.2%, 18.7%, and 10% decline in Pn, Gs, and Tr, respectively, but did not negatively affect fruiting branches or individual boll weight. The DSS demonstrated its capacity to maintain leaf photosynthesis and growth of cotton under arid conditions with better water use efficiency.