The reference genotoxic agents 2-aminoanthracene (a metabolism dependent weak clastogen), 5-fluorouracil (a nucleoside analogue, characterised by a steep dose response profile), colchicine (an aneugen that inhibits tubulin polymerisation), benzo[a]pyrene (a polycyclic aromatic hydrocarbon requiring metabolic activation), cadmium chloride (an inorganic carcinogen), and cytosine arabinoside (a nucleoside analogue that inhibits the gap-filling step of excision repair) were tested in the in vitro micronucleus assay using the Chinese hamster ovary (CHO) cell line at Covance Laboratories, Harrogate, UK. All chemicals were treated in the absence and presence of cytokinesis block (via addition of cytochalasin B) with this work forming part of a collaborative evaluation of the toxicity measures recommended in the draft OECD Test Guideline 487 on the In vitro Mammalian Cell Micronucleus Test (MNvit). The toxicity measures used, detecting a possible combination of both cytostasis and cell death (though not cell death directly), were relative population doubling, relative increase in cell counts and relative cell counts for treatments in the absence of cytokinesis block, and replication index in the presence of cytokinesis block. All of the chemicals tested either gave marked positive increases in the percentage of micronucleated cells with and without cytokinesis block, or did not induce micronuclei at concentrations giving approximately 50-60% toxicity (cytostasis and cell death) or less by all of the toxicity measures used. The outcome from this series of tests supports the use of relative increase in cell counts and relative population doubling, as well as relative cell counts, as appropriate measures of cytotoxicity for the non-cytokinesis blocked in vitro micronucleus assay.