This retrospective study examines the role of periodontal probing depth in clinical decision-making. The expected values of no treatment and surgical and non-surgical therapies were obtained by combining the probability of each treatment outcome and utility values denoting the preference for each outcome. The expected value for non-surgical therapy was higher for sextants exhibiting a positive response to initial therapy than sextants which did not respond to initial therapy (0.735 versus 0.706). This trend was not observed for sextants treated surgically. Surgical therapy was effective over all levels of disease severity and was the preferred form of therapy with respect to reduction of probing depth except for sextants exhibiting 4 to 5 mm pockets. In this latter situation, the expected value at 3 years for non-surgical and surgical therapy was 0.795 and 0.792, respectively. Neither form of therapy could consistently achieve periodontal probing depths equal to or less than 3 mm throughout a given sextant. This study facilitates the selection of an optimal therapeutic strategy with respect to periodontal probing depth.