One hundred and fourteen patients with a medical history of recurrent upper respiratory tract infections (at least four episodes in the year preceding this open comparative study) were randomly assigned to three groups of 38 patients each. The first group received, by sublingual route, Ismigen, a new immunostimulating lysate (Polivalent Mechanical Bacterial Lysate, PMBL) obtained by mechanical lysis of 48 billion bacteria commonly responsible for upper respiratory tract infections; the second group received an oral immunostimulating lysate (CLBL) obtained by chemical lysis of 36 billion bacteria. A third group served as control and did not receive any immunostimulating treatment (Control NT). One tablet a day of PMBL was given to the first group for the first ten days during three consecutive months; the patients of the second group received one capsule a day of CLBL with the same treatment schedule. At the end of the treatment period the patients of the 3 groups were followed up for 3 months. The primary end point was the number of acute upper respiratory tract infections (URTIs) that occurred during the three months of treatment and three months of follow-up. Secondary endpoints were: the number of patients free from disease, the duration of infectious episodes, the number of working days lost because of disease, the need for a concomitant antibiotic treatment, and the safety of the two treatments. During the treatment period the mean number (+/- SD) of URTIs per patient was 0.34 (0.48) in the PMBL group, 1.0 (0.83) and 1.23 (0.77) in the CLBL and Control NT groups, respectively. Results of PMBL treatment were significantly better (p < 0.05) than the results in the other two groups; CLBL was not significantly different from the control group. In the three months of follow-up, the mean number (+/- SD) of URTIs per patient was: 0.42 (0.55) in the PMBL group, 0.92 (0.67) in the CLBL group, and 1.55 (0.88) in the Control NT group. The PMBL group was significantly better than the other two (p < 0.05). During the 6-month study, significantly more patients of the PMBL group remained free from respiratory infections in comparison to the other two groups. As regards other secondary end points (duration of infectious episodes and number of working days lost), the mean values of the PMBL group were statistically significantly lower than those of the other two groups, in both the treatment and follow-up periods. No PMBL treated patient needed concomitant administration of an antibiotic, while 9 patients of the CLBL group received such treatment (p < 0.05). No patient reported adverse events.