Small-scale laboratory tests for sizing and designing disc filters are typically performed by using leaf test equipment. Although the basic tests are fairly simple and quick, the accuracy and repeatability of the results may be poor due to various reasons. In addition to errors caused by variations in the structure of the equipment, also the skills and experience of the person performing the tests is of great importance. This paper introduces an experimental study carried out for defining the most important sources of errors in these kinds of laboratory tests and for estimating their influence on the final test results. The results obtained in this study were utilized for designing an improved version of the test equipment and also for creating a standard procedure for performing the tests. The results acquired with the small-scale laboratory tests were in good agreement with the performance results obtained from full-scale ceramic disc filters operating in industrial processes. (C) 2011 Elsevier Ltd. All rights reserved.