A double-blind random selection comparison was made of the therapeutic effects in acute herpes zoster of 40% idoxuridine (IDU) dissolved in dimethyl sulphoxide (DMSO) compared with DMSO and saline flavoured with garlic. Thoracic (80 patients) and trigeminal (42 patients) zoster were investigated separately. The patients were evaluated daily until skin healing and then at 1, 3 and 6 months by registering pain, paraesthesia and sensitivity disturbances as well as by clinical and photographic evaluation of the skin lesions. Duration of pain was positively correlated to age, to delayed healing and to elevated temperature in the acute phase of zoster. The period of pain before skin eruption was considerably longer in thoracic than in trigeminal zoster, while the latter was associated with a more severe inflammatory reaction, more neurologic sequelae, but also by a faster healing of the skin lesions. IDU was highly effective in shortening the period of pain and improving skin healing in trigeminal zoster, while no effect of IDU was observed in thoracic zoster. The reason for this difference is presently not understood.