This prospective multicenter intra-individual randomized controlled clinical trial was designed to compare the efficacy of guided tissue regeneration (GTR) with bioresorbable barrier membranes versus access flap surgery, in intrabony defects. 2 similar defects were selected in each of 23 patients and randomly assigned to 1 of the 2 treatments. Surgery consisted of an identical procedure except for the omission of the barrier membrane in the flap control sites. At 1-year, probing pocket depth reductions were 4.3+/-2.3 mm in GTR treated sites and 3.0+/-1.5 mm in the flap control sites (p=0.02, paired t-test). Clinical attachment level (CAL) gains were 3.0+/-1.7 mm in the GTR sites and 1.6+/-1.8 mm in the control sites (p=0.009, paired t-test). A subset analysis, performed according to the initial depth of the intrabony component of the defects (INFRA), indicated that in shallow defects (INFRA < or =3 mm) treated with the access flap alone, CAL gains were 1+/-1.5 mm, while in deep ones (INFRA > or =4 mm) they were consistently greater (1.9+/-1.9 mm). The % CAL gains, calculated as the % of the baseline intrabony component depth, however, were almost identical in the 2 subpopulations (45.8+/-64.7% in shallow and 43.8+/-37.6% in deep defects). Similarly, in the GTR sites, linear CAL gains were greater in deep (3.7+/-1.7 mm) than in shallow defects (2.2+/-1.3 mm), but no differences were observed in terms of % CAL gains (76.7+/-27.7% and 75.8+/-45%, respectively). The frequency distribution of CAL changes expressed as %s of the baseline INFRA indicates that most of the sites treated with GTR (73% in shallow and 92% in deep defects) gained 50% or more CAL. Furthermore, many defects (64% of shallow and 33% of deep defects) reached 100% of CAL gain. The present study demonstrated that: (i) GTR with bioresorbable barrier membranes resulted in a significant added benefit in comparison with access flap alone; (ii) the linear amounts of CAL gains were greater in deep than in shallow defects; (iii) CAL gains expressed as %s of the baseline depths of the intrabony component, were similar in shallow and deep defects; (iii) the regenerative procedure tested in the present study resulted in CAL gains equal to the depth of the intrabony component of the defect in some, but not in most of the instances.