Abstract: Water retention at −1500 kPa is a crucial parameter in soil survey interpretations and in many hydrologic, agronomic, and ecological applications. Because it is impractical to measure −1500 kPa water contents everywhere in a soil survey area, it is often estimated using a model that is considered rule of thumb. The objectives are to validate this model and determine if a superior model can be developed using general linear models. Validation results indicate a consistent bias between the measured and predicted values. The 1500 kPa model was found to overestimate the contribution of organic matter, and the 1500 kPa water-to-clay ratio (WCR) of 0.4 was shown to be in agreement with measured ratios from a wide range of soils. For soils dominated by specific clay minerals, the 1500 WCR ratios varied from 0.25 for soils containing gypsum to 0.50 for soil clays containing hydroxy-interlayered vermiculite. The accuracy of the model could be improved if the clay mineralogy is known and the corresponding 1500 WCR is used instead of 0.4. The development of a new model resulted in total clay, organic carbon, and cation exchange capacity explaining 91% of the variation in water contents retained at −1500 kPa. The new model had the lowest root mean square error of 1.967, indicating a more accurate model than just using total clay or total clay and organic carbon as predictor variables. On a wide range of soils, validation of the newly developed model indicated no bias and was superior to the rule of thumb model in accuracy. The new model will be useful in soil survey as an approximation when data are lacking and is not intended to replace measured values.