10 min, air dried, and dampened in deionized water. Tissue was scraped off the slides, suspended in 50-100 lL of a solution containing 0AE1 mol L )1 Tris-HCl (pH 8AE0) and 1 lg lL )1 proteinase K, incubated at 55°C overnight, boiled for 10 min, and finally stored at )20°C until analysed. To amplify HSV DNA using PCR, a pair of primers (Research Genetics, Huntsville, AL, U.S.A.) defining a 92-bp segment of the DNA polymerase gene was used, as described by Cao et al. 10 PCR reactions were performed with AmpliTaq DNA polymerase (Applied Biosystems, Foster City, CA, U.S.A.) under the following conditions: denaturation at 94°C for 45 s, annealing at 63°C for 30 s, and elongation at 72°C for 30 s. After 40 cycles of amplification, the reaction mixture was electrophoresed on a 2AE5% agarose gel (FMC, Rockland, ME, U.S.A.) and stained with ethidium bromide to visualize the DNA. Southern blot hybridization was performed with an internal 32 P-labelled HSVspecific oligonucleotide probe and was evaluated on X-ray film.All 24 PLE specimens that we examined were negative for HSV DNA by PCR analysis and subsequent Southern blot hybridization (Table 1, Fig. 1). In contrast, 10 of 31 EM specimens (32%) were positive for HSV DNA. This rate of HSV positivity in EM cases was comparable with 35-72% positivity rates seen in previous studies. 1,4,5 Interestingly, Yokoi et al. 4 detected HSV DNA in four of seven specimens from postherpetic EM but in none of three specimens from photoinduced EM. However, our inability to find HSV DNA presenting in any PLE skin samples contradicts the hypothesis that a direct immune response to HSV in the skin is involved in the pathogenesis of PLE.