Accurate estimation of crop yield under aeration stress is crucial for field water table management. In this study, the CROPR crop model was improved in two aspects: (i) a new aeration factor, which was related to a drainage index, was proposed and used to represent the condition of soil aeration; and (ii) a multiplicative structure, instead of the original additive structure, was used in the calculation of dry matter accumulation to include the after-effect of aeration stress. Four-year lysimeter experiments on cotton (Gossypium hirsutum L.) growth under aeration stress were conducted from 2008 to 2011 to calibrate and validate both the original and improved CROPR. The results indicated that the improved CROPR performed better than the original CROPR and was suitable for simulating cotton yield under aeration stress. In the calibration, with the improved CROPR, the root-mean-squared error (RMSE) of seed cotton yield was 832.84 kg ha^–1 with a normalised value (NRMSE) of 15.87%, whereas with the original CROPR, the RMSE was 973.03 kg ha^–1 with an NRMSE of 18.55%. In the validation, with the improved CROPR, the RMSE of seed cotton yield was 686.22 kg ha^–1 with an NRMSE of 14.87%; with the original CROPR, the RMSE was 1019.02 kg ha^–1 with an NRMSE of 22.08%.