The efficacy of early versus late treatment with acyclovir and valaciclovir on zoster-associated pain was assessed from two databases (1076 patients) that were compiled from randomized trials. Early treatment was started < 48 h and late treatment was started 48-72 h after the onset of cutaneous herpes zoster. Median times to complete resolution of zoster-associated pain were 28 and 62 days, respectively, for patients (> or = 18 years of age) treated with acyclovir and placebo within 48 h (hazard ratio [HR], 1.68; 95% confidence limit [95% CL], 1.19, 2.38) and 28 and 58 days, respectively, for those treated later (HR, 2.20; 95% CL, 1.03, 4.71). In the valaciclovir versus acyclovir study (in patients > or = 50 years of age), the corresponding figures were 44 and 51 days for patients treated early (HR, 1.28; 95% CL, 1.03, 1.60) and 36 and 48 days for those treated later (HR, 1.40; 95% CL, 1.04, 1.87). Acyclovir significantly shortened the time to complete resolution of zoster-associated pain compared with placebo (and valaciclovir was superior to acyclovir in this regard) even when therapy was delayed up to 72 h after rash onset.