A theoretical but as yet not conclusively proved risk of chloramphenicol induced idiosyncratic aplastic anaemia exists with topical ophthalmic therapy, with the absolute, but highly improbable, maximum risk of death (equalling that of systemic therapy) being 1 in 50,000 to 90,000. To put this in realistic perspective, one must note that the comparable risk of fatal anaphylaxis resulting from penicillin therapy, from any route, is similar at 1 in 100,000. Indeed, it has been noted recently that with more than 200 million ocular chloramphenicol products dispensed in the UK in the past 10 years, only 11 reports (all non-fatal) of suspected topical chloramphenicol induced blood dyscrasia have been reported to the Committee on the Safety of Medicines since 1966. One also has to consider that inadvertent exposure to minute quantities of chloramphenicol (ng/ml) may occur through consumption of livestock that have been treated with chloramphenicol. Broad statements condemning topical chloramphenicol need to be tempered with its proved safety, tolerance, cost, and efficacy while acknowledging an extremely remote risk of the very serious adverse effect of drug induced aplastic anaemia. Risk-benefit assessment is the duty of all prescribing physicians and a decision to prescribe or not prescribe must be made on the basis of personal judgment and an awareness of the statistics in perspective. The only known factor to be associated with vulnerability in the case of topical chloramphenicol is family history. There is no evidence to date that suggests children are any more susceptible than adults.