In contrast to existing similar-media scaling methods which minimize sum of squared
differences (SS) between a mean soil water retention curve and scaled soil water pressure data, we propose a new method involving maximization of the correlation coefficient (R) between measured and estimated soil water pressure heads. With this new criterion,
multivariate statistical procedures are implemented resulting in an explicit non-iterative solution for the set of scale factors describing the spatial variability of measured soil water retention curves. Performance of the proposed method was tested with published data of insitu soil water retention measurements made at a site in North Dakota, USA. Scaling was successful as indicated by substantial reduction in SS and matched reported performances of existing iterative methods. For the dataset used, our algorithm provides the optimal solution in a non-iterative manner and introduces minimal distortions into original retention measurements.