In a randomised, double-blind, controlled study of the effect of prednisolone on the development of post-herpetic neuralgia 78 patients with herpes zoster whose pain and exanthema had been present for less than 96 h were given 800 mg acyclovir five times daily for 7 days and prednisolone in a total dose of 575 mg, starting with 40 mg daily in the first week and tapering off over the next 2 weeks. 18 (23%) of the patients had post-herpetic neuralgia at 6 months after the acute zoster, 9 (24.3%) having received prednisolone and 9 (22.5%) placebo. The 95% CI for the difference between the placebo and prednisolone groups in the proportion of patients having pain at 6 months was minus 17% to plus 20%. Prednisolone, however, relieved pain for the first 3 days. The 1-2 week interval between admission and reappearance of pain and development of triggered pain seems to be the time needed to establish neuralgia. Once established, the type and intensity of pain remained largely unaltered.