The Silt Density Index is applied world-wide for many decades to determine the fouling potential of feed water of reverse osmosis systems and more recently to judge the performance of micro- and ultrafiltration systems. However there are growing doubts about the reproducibility and accuracy of this test. Currently, the Silt Density Index (SDI) is applied without any correction for temperature, applied pressure and membrane resistance. Besides that, the SDI is not based on any fouling mechanism which affects its reproducibility and accuracy. To identify opportunities for improvements, existing mathematical fouling models were further extended to study the effect of temperature, applied pressure and membrane resistance on the SDI value under four different fouling mechanisms. Significant variations in SDI values are observed mathematically as a result of differences in temperature and membrane resistance for the same water quality. The fouling mechanisms are described by the relationship between the filtrated volume w and the total resistance R. The sensitivity of the SDI for variations in the testing parameters theoretically increases when the relation between w and R is stronger. The SDI increases with an increase in feed temperature and applied pressure. Temperature has a substantial effect on SDI. As a consequence it is not recommended to compare SDI values measured at different temperatures. The SDI value decreases when membranes with a high resistance are used. To achieve a more reliable SDI, the use of a standardized membrane with constant properties, in particular having a narrow resistance range, is recommended. (C) 2011 Elsevier B.V. All rights reserved.