Gingival crevicular fluid (GCF) was collected from two healthy, two gingivitis and two periodontitis sites of two groups of individuals presenting for treatment of chronic adult periodontitis (group 1, 25 subjects; group 2, seven subjects) and from distal approximal sites of two incisors and one molar of 10 subjects with periodontal health. GCF eluates of periodontitis group 1 and controls, prepared by a technique that lysed polymorphonuclear leucocytes (PMN) in the samples, were assayed for functional neutrophil elastase (NE) and immunoreactive alpha 1-antitrypsin (alpha 1-AT) and alpha 1-antitrypsin-neutrophil elastase complex (alpha 1-AT-NE). Periodontitis group 2 GCF eluates, generated by a method that did not disrupt PMNs, were assayed for functional NE in the presence and absence of a specific NE inhibitor. A greater amount of NE (ng/5-s sample) was found in eluates of GCF from diseased sites irrespective of whether or not the eluates contained products of lysed PMNs. However, the GCF eluates prepared without disrupting PMNs contained only about one-tenth as much NE as eluates of corresponding sites that included constituents of lysed PMNs. The amount of alpha-AT in GCF was insufficient to inactivate most of the NE available for release into the gingival sulcus at either healthy or diseased sites. In addition, much of the alpha 1-AT in GCF was not complexed with NE under conditions of excess NE. More than 90% of the NE in GCF from each site category was inactivated by the NE specific inhibitor. It is concluded, because of the large quantity of NE available in PMNs compared to the amount of NE inhibitors in GCF, that at least locally transient free NE occurs, which contributes to tissue destruction in chronic adult periodontitis.