A set of simple test cases have been defined, within the activities of the CIE technical committee 3.33 and Subtask C of the IEA SHC Task 31, based on analytical or experimental references, with the objective of assessing lighting computer programs. The present paper aims to show the applicability and the usefulness of these test cases through the analyses of a conducted set of benchmarks. It is shown that the use of the CIE test cases allows to verify the level of accuracy of the tested programs with respect to the lighting physical laws that are highlighted separately within the different test cases. The comparison of the results of the different tested programs shows also that a program with a high accuracy in respect to a given aspect of lighting simulation is not necessarily accurate in simulating other aspects, which proves the usefulness of the CIE test cases in identifying the domain of applicability of a tested program. (c) 2005 Elsevier Ltd. All rights reserved.