Eighty-four patients with bacterial vaginosis were examined in an open randomized trial, the aim of which was to define clinical results and the microbiological panorama after topical treatment for 1 week with either an acetic acid jelly (A), an estrogen cream (B), a fermented milk product (C) or metronidazole (D). After exclusion because of chlamydia infection (15 cases) or for other reasons, 61 cases remained for complete evaluation 4 weeks after the start of treatment. Clinical cure was obtained in 3 cases out of 17 on regimen A, in 1 out of 16 on regimen B, in 1 of 14 on regimen C, and in 13 out of 14 on regimen D. The patients were conclusively either symptomless or symptomatic when examined on 113 occasions. Statistically significant reduction after treatment resulting in relief of symptoms was observed in the numbers of corynebacteria and anaerobic cocci, whereas lactobacilli increased in numbers. The instillation of high numbers of Lactobacillus acidophilus (C) into the vagina cured only 1 patient and did not influence the predominance of lactobacilli in the vagina at the follow-up examination. The difference in microbiological profile of women in symptomatic and asymptomatic conditions becomes still more apparent when the results of the present and previously published studies on the subject by the present group of investigators are combined. The symptomatic woman is significantly more often harboring corynebacteria, Gardnerella vaginalis, peptostreptococci, peptococci, eubacteria and Bacteroides species. Lactobacilli are significantly reduced in numbers. However, only 51 % of our previously symptomatic, but now symptomless women show predominant growth of lactobacilli, which is less than expected for healthy women. Therefore, lactobacilli may not be the specific factor that protects against bacterial vaginosis.