ABSTRACT: (see part I) The study initiated with 69 patients, out of which, 55 were available at the completion of the trial. Methods: evaluations were made at the beginning 3, 6, 12 and 24 months. The initial exam included the following criteria: socio demographic (age, gender, tobacco use and family grouping). The clinical criteria included were: probing depth, clinical attachment, superficial bleeding, probing bleeding and sub gingival micro flora. Microbiological samples were taken from 6 different sites at the beginning, 12 and 18 months. The statistical analysis included the Chi square and Student’s T test for paired averages in time, both with a level of significance of 5%. Also a logistic regression analysis was done to find predictive variables of response to treatment. The purposes of this study were to determine the microbial composition of periodontal pockets, and to evaluate the response of non responsive patients to a second alternative therapy with systemic antimicrobials, two years after its application and comparison with the responsive patients. Results: three months after treatment the changes (loss or gain) in clinical attachment (CA) allowed us to classify the patients as: Responsive (R) to 57 patients (82.6%) and Non Responsive (NR) to 12 (17.4%). The variables analyzed did not show individual or group association with the type of response to treatment, with the exception of supra gingival index plaque > 2 (P = 0.05). Conclusions: follow up after two years indicated that the alternative treatment was beneficial for the NR that behaved during this period in the same way as the R, both showing stability in the CA. The prevalent bacteria were Porphyromona gingivalis (P .g.), Porphyromona sp (P sp.) and Prevotella Intermedius (P i.). 45% of the responsive patients showed enteric bacteria while the non responsive had 78%.