The purpose of this study was to determine the clinical response to local delivery of tetracycline in relation to clinical and microbiological conditions of the other teeth. 4 deep pockets were monitored in 19 subjects with multiple deep periodontal lesions and high counts of P. gingivalis. In 9 patients (LT) only 2 of the selected lesions were treated by placement of tetracycline fibers (Actisite), while the rest of the dentition was left untreated. In the other 10 patients, all teeth were supragingivally scaled and then treated by application of polymeric tetracycline HCl containing fibers, the whole dentition was subject to full mouth scaling and root planing, and the patients rinsed with 0.2% chlorhexidine (FT). A significant reduction in mean PPD was observed in all treated sites after two months. This reduction was maintained over the following 4 months. The magnitude of the effect was significantly greater in the FT group (1.74 mm) than in the LT group (0.88 mm). The mean attachment level changes were similar after 2 months in locally and fully treated subjects. A tendency of relapse was noted for treated sites in LT patients from month 2 to 6. A level of statistical significance was not reached for this effect. Data from measurements recorded at 6 sites around all teeth in the full mouth treated patients were analyzed using multiple linear regression. This analysis showed local changes in PPD and AL were significantly and strongly correlated with the baseline value of the respective parameter at the same site. In addition, more pocket depth reduction was noted if a site was not bleeding on probing at 6 months, if the location of a site was not approximal and if the tooth was not a second molar. Sites located on second molars showed also less AL gain than sites located on other teeth. Smokers showed significantly less reduction in PPD and significantly less AL gain. Furthermore, if subjects had a high % of pockets deeper than 4 mm at baseline they showed significantly less attachment gain.