Intracrevicular antimicrobial therapy is consistent with the site-specific nature of periodontitis. Considerable research has focused on the use of nonresorbable fibers. However, a bioerodible system is desirable. The purpose of this study was to assess tetracycline release and safety following a single application of a syringable 35% tetracycline hydrochloride in a lactic-glycolic acid gel. 31 generally healthy adult volunteers (mean age = 59 years) were enrolled in and completed this randomized, double-blind eight day study. 2, 6-10 mm non-adjacent interproximal pockets that bled on pocket probing were chosen as experimental sites in each subject. I experimental site and the surrounding gingival crevice received small particle size tetracycline in gel while the other site received larger particle size tetracycline in gel. Gingival crevicular fluid (GCF) was collected prior to treatment and 15 min, 1, 2, 3, 4 and 8 days post-treatment. GCF tetracyline concentrations were determined by agar diffusion bioassay and GCF volume measurements. 61% and 71% of sites had > or = 100 micrograms/ml tetracycline 3 days following application of large (mean concentration = 430 +/- 92 micrograms/ml) and small particle gels (mean concentration = 418 +/- 70 micrograms/ml), respectively. 37% and 55% of sites had measurable tetracycline 8 days after placement of large (mean concentration = 86 +/- 31 micrograms/ml) and small particle gels (mean concentration = 293 +/- 79 micrograms/ml), respectively. The most common adverse event was "bitter taste" (10% of subjects). Based upon the reduction in probing depths and % of sites bleeding on probing at 8 days relative to pretreatment, and the absence of any serious adverse events, it is concluded that these bioerodible gels are safe, and since the bacteriostatic range for most putative periodontopathogens is in the 2-10 micrograms/ml range, the tetracycline levels observed at days 3 and 8 likely represent significant antimicrobial efficacy.