There is a need for information on the validity of probing depth and attachment level measurements made with automated probes. In this study, 34 teeth in 9 patients were measured with the Florida probe from points marked with a bur prior to extraction. After extraction, connective tissue attachment levels were measured from the same points with a dissecting microscope. Mean measurements of attachment level were similar with probing (5.13 +/- 2.08 mm) and laboratory (5.18 +/- 2.26 mm) assessments. The correlation between these measurements was 0.72 (p < 0.001), and a t-test of the paired measurements showed no difference (t = 0.44; p = 0.66). The Florida probe thus showed satisfactory validity for this group of measurements of advanced periodontitis. However, there was substantial lack of agreement between individual probing measurements and the validity criterion of laboratory attachment level measurements, and the intra-class correlation coefficient was 0.46. It was concluded that the Florida probe was suitable for studies in which measurements were averaged, but not for longitudinal study of individual sites.