In order to cope with large amounts of samples for chemiluminescence (CL), vials were replaced with microplates. Although various types of plates have been commercially available for quite some time and the free-plate mode is advocated by the producer of the counter, little is known about their impact on the outcome of CL measurements. We tested two different 24-well microplates and six different 96-well microplates in two different luminometers, and results were compared with those achieved with vials. Before these comparative tests, we attempted to optimize measurement conditions. CL sensitivity was highest with luminol concentrations of 0.8–3.3 µmol/L, PMA concentrations of 0.06–80 µmol/L, a pH value of 10 and a temperature of 20°C. An indirect correlation was found between fluid volume and yield in counts: the lower the volume, the higher the counts. With regard to sensitivity and cross-talk, the 96-well Isoplate™ was superior to all other plates tested. While all white plates tested gave acceptable results, usage of the black 96-well plates resulted in an extremely low sensitivity. Plates designed for cell culturing gave even lower counts and a cross-talk of up to 31%. All attempts to reduce cross-talk and improve sensitivity, such as aluminium foil or grids, irrespective of the position of the photomultiplier, did not give results comparable to the original 96-well isoplate. Our results suggest that, with the exception of black 96-well microplates and cell culture plates, all other plates tested have a sufficient sensitivity when compared to vials and acceptable cross-talk, the 96-well Isoplate™ being the best. Both types of luminometers used gave reproducible results, Wallac having a somewhat higher sensitivity, Canberra Packard somewhat less cross-talk. Copyright © 1999 John Wiley & Sons, Ltd.