In connection with genital infections observed in private practice among women, the collection of 1,836 cervicovaginal samples in 1987 and 368 in 2001 enabled clinical and biological comparisons through the analysis of the cervicovaginal flora. First of all, a diagnosis of STI (sexually transmitted infection) is rarely made. However, in the majority of cases, this examination allows either for the precise treatment of a confirmed infection (chlamydia, trichomoniasis, candidiasis, gonorrhea, bacterial vaginosis) or helps in reassessing prolonged treatment that is often unnecessary or even poorly tolerated. Not all cases of vulvar pruritus, for example, with or without pelvic pain, are by any means due to a yeast infection. It has become clear to us that attempting to base curative treatment on isolated or associated clinical symptoms is futile because, while they may sometimes suggest a cause, they have only a low positive predictive value. (The PPV of the pruritus+pelvic pain combination is only 10% for chlamydia and 45% for candidiasis.) The diagnosis of vaginosis, proposed over the past decade to improve the diagnosis of vulvovaginitis, could only be made in 13% of cases. The only difference noted between our two studies was the lower frequency of isolates of Neisseria gonorrhoeae, Chlamydia, and Ureaplasma in 2001, with all other conditions remaining identical, except for a smaller number of patients.