Between about 1955 and 1970, some 100,000 Japanese were diagnosed as having subacute myelooptic neuropathy (SMON), a new disease characterized by abdominal and neurological manifestations, the former nearly always preceding the latter. Circumstantial evidence obtained in 1969-70 suggested that SMON might have been caused by clioquinol (CQL), a gastrointestinal disinfectant, and led to the suspension of further sales of CQL in Japan. However, several inconsistencies for the CQL theory of SMON have now emerged; first, CQL had been widely used in Japan for nearly 20 years before SMON occurred. Secondly, the SMON epidemic began to subside several months before CQL sales were suspended. Thirdly, a large proportion of SMON patients--probably about one-third and possibly more--had not taken CQL within six months of the onset of the disease (the modal interval between first taking CQL and the onset of SMON being about three weeks, and more than 100 days in only 4% of SMON patients); of the remaining two-thirds or so, many had taken CQL as part of the treatment of the first (that is, abdominal) symptoms of SMON itself. Fourthly, there was no dose-response relationship. Finally, SMON rarely, if ever, occurred outside Japan. CQL could, however, have been involved in the causation of SMON as an optional enhancer of some other necessary cause; the history of post-war environmental pollution in Japan is compatible with this hypothesis. Over-readiness to accept postulated toxic effects of medicines and chemicals as proven is likely to do at least as much harm as good to individual and community health.