The aim of this study was to evaluate the tuberculin skin test in individuals vaccinated with bacillus Calmette-Guérin (BCG) using 2 IU of RT-23. One hundred ninety-six individuals aged 22-40 years-old who had been vaccinated with BCG between 1965 and 1974 were enrolled along with 375 non-vaccinated individuals of the same age and with similar level of risk of infection. The positive predictive value of the test was assessed for three levels of response as indicated by areas of thickening in three diameters: 5, 10 and 15 mm. Vaccinated individuals with negative results were given a second skin test 7 days later to detect a booster effect. Positive diameters 5 mm were observed in 66% of the vaccinated individuals and 24% of the non-vaccinated subjects. Positive diameters 10 mm were observed in 51% of the vaccinated individuals and 19% of the non vaccinated ones. Positive diameters 15 mm were observed in 29% of the vaccinated subjects and in 13% of the non vaccinated ones. The differences were significant for all diameters. The positive predictive value of the test was 36.4% for a diameter 5 mm, 37.6% for diameter 10 mm and 44.8% for diameter 15 mm. The booster effect was detected in 25.8% of the vaccinated individuals who had tested negative at first. In vaccinated individuals, no guidelines can be established to guarantee that a positive reaction is due to infection by Mycobacterium tuberculosis infection, although the likelihood of infection (increased positive predictive value) increases with diameter. It is also impossible to fix a time limit. A second skin test is needed to detect a booster effect in all vaccinated individuals whose first test is negative.