Topical locally delivered minocycline is an adjunctive to non-surgical periodontal treatment, but there are few reported trials. Previous trials have reported differences between changes in probing depth in treatment and control groups, but no differences in probing attachment level. In the present study, 30 subjects were paired according to gender, age, ethnic group, smoking habits, and probing depths. Both groups received intensive oral hygiene education and root planing with local anaesthesia. Active or placebo gel was placed subgingivally at planed sites in each subject according to a double-blind protocol, immediately after instrumentation, and 2 and 4 weeks later. A periodontal examination was made with a constant force probe before instrumentation, and 6 and 12 weeks later, 2 subjects failed to complete the study, and their pairs were therefore not included in the analysis. Results were tested with analysis of covariance. Differences between groups in mean probing depth did not reach statistical significance at any visit (baseline: test (T) = 5.93 mm, control (C) = 5.74 mm; 6 weeks: T = 3.53 mm, C = 3.63 mm; 12 weeks: T = 3.29 mm, C = 3.44 mm), but mean probing attachment levels were different (p < 0.05) at both reassessments (baseline: T = 6.86 mm, C = 6.83 mm; 6 weeks: T = 4.93 mm, C = 5.30 mm; 12 weeks T = 4.91 mm, C = 5.27 mm). There was also a difference in the number of sites with bleeding on deep probing at 12 weeks (p < 0.05). This trial showed that adjunctive minocycline gel provided a more advantageous outcome for nonsurgical periodontal treatment in terms of probing attachment level and bleeding on deep probing. This trial was a good example of experimental, as opposed to community, design, and used limited resources to show a clear result.