BACKGROUND AND OBJECTIVE: In spite of four decades of studies on gingival crevicular fluid, no data have been reported on the repeatability of gingival crevicular fluid collection and the subsequent quantification procedures. The present study reports, for the first time, on the repeatability and method error of gingival crevicular fluid collection and quantification, as determined through its alkaline phosphatase (ALP) activity. Diagnostic considerations are then explored. MATERIAL AND METHODS: Twenty-seven healthy subjects (17 women and 10 men; mean age ± SD, 21.2 ± 4.8 years) with optimal periodontal status were enrolled according to a blind prospective design. The gingival crevicular fluid was collected at baseline, and after 1 d, 1 wk and 3 mo. At each clinical session, two consecutive rounds of gingival crevicular fluid collection were made from each of the four maxillary incisors, allowing the recovery of resting and flow gingival crevicular fluid. The total ALP activities were determined spectrophotometrically, and repeatability and method errors for the resting, flow and overall (resting + flow) gingival crevicular fluid ALP activities were calculated, relative to the corresponding baseline levels. RESULTS: No significant differences were seen over time, although the flow gingival crevicular fluid ALP activity was generally lower than that for the resting gingival crevicular fluid. The method errors ranged from 40 to 58%, with the flow and overall gingival crevicular fluid activities showing the highest and lowest errors, respectively. CONCLUSION: Reliable use of the gingival crevicular fluid ALP collection and quantification, both in research and diagnosis on an individual basis, should take into account relevant errors, and variations are to be considered as true only above relevant thresholds.