Nonoxynol-9 (N-9) is virucidal in vitro, and is therefore a candidate microbicide for preventing sexual transmission of HIV. However, the activity of N-9 is nonspecific, suggesting that virucidal levels may produce adverse effects including epithelial disruption, inflammation of the genital mucosa, or both. A randomized placebo controlled trial of daily use of 100 mg of N-9 took place for 1 week in 40 female volunteers. Outcome measures included symptoms, colposcopic and histologic changes in the genital tract, and impact on vaginal flora. Genital irritation was reported by 10 of the N-9 and 5 of the placebo group. Colposcopy showed erythema in 9 of the N-9 group and 2 of the placebo group. Histologic inflammation was found in 7 of the N-9 group and 2 of the placebo group. Inflammatory changes were characterized by patchy infiltration of the lamina propria predominantly with CD8+ lymphocytes and macrophages, in the absence of epithelial disruption. A transient reduction in numbers of lactobacilli was observed in 9 of the 15 women using N-9, and 6 of 18 women using placebo. N-9 used for 7 days in a standard spermicidal dose was associated with increased irritation, colposcopic and histologic evidence of inflammation and was more frequently associated with reduction in numbers of lactobacilli during gel use. The clinical significance of the recruitment of cells susceptible to HIV infection to the genital mucosa is unknown but raises concerns about the suitability of N-9 as a microbicide when given in this dose.