In order to standardize and assess the quality of immunophenotyping of leukaemias and lymphomas for diagnostic purposes, a cooperative study group in the Netherlands, SIHON, has formulated guidelines for the composition of antibody panels to be applied and guidelines for the interpretation of the marker analysis. To assess the value of these guidelines frozen cell samples of three patients with different haematological malignancies were sent to the 26 participating laboratories twice a year. Here we present the results with respect to the marker analysis and to the immunological diagnosis on 387 samples from 18 patients. A large inter-laboratory variation was seen in the percentage of positive cells for each marker, which influenced the valuation of a marker to be discordant positive in up to 23% and discordant negative in up to 40%. No single major factor could be traced to explain the large variation in the results. However, probably due to the balanced composition of the antibody panel and to the application of the guidelines for interpretation, this variation did not much influence the agreement in immunological diagnosis. In only 13/387 samples (3.3%) differences in the percentage of positive cells caused disagreement in the final diagnosis. In 23 samples (5.9%) the disagreement was due to an incorrect application of the guidelines. Quantitative data of single observations obtained from different laboratories, in which the materials and methods are not standardized, cannot be compared; but standardization of guidelines for marker sets and for interpretation contributes to a high grade of agreement in immunological diagnosis.