Inefficient water and nitrogen management significantly constrains the yield potential of Sea Island cotton in arid regions. Field experiments in southern Xinjiang, China (2022–2023) evaluated the interactive effects of irrigation frequency (F₅, F₇, and F₉: irrigation every 5, 7, and 9 days) and nitrogen rates (N₀, N₇₀, N₈₅, and N₁₀₀: 0, 289.8, 351.9, and 414 kg ha⁻¹) on cotton productivity, fiber quality, and resource use efficiency using a split-plot design. Partial least squares path modeling (PLS-PM) and ternary trade-off analysis were integrated to quantify yield formation pathways and the multi-objective balance among water productivity (WP), partial factor productivity of nitrogen (PFPN), and economic benefit (EB). The F₅N₈₅ combination increased seed cotton yield by 7.26%-8.77% over the conventional benchmark while maintaining premium fiber quality. PLS-PM indicated that reproductive nitrogen accumulation at the boll opening stage (path coefficient = 0.43) was the strongest statistical associate of yield, suggesting that late-stage nitrogen partitioning to reproductive organs is a more proximate yield correlate than vegetative biomass. Ternary evaluation coupled with TOPSIS confirmed that F₅N₈₅ achieved the most balanced multi-objective performance across both seasons, avoiding the severe efficiency trade-offs observed under low-frequency irrigation (F₉).