The handpiece of a Florida sleeve probe was modified to create a flange with the capability to detect the cemento-enamel junction (CEJ). This new instrument (the Pressure-controlled, Automated, Standardised Handpiece or Florida PASHA probe) was used to determine whether (a) the CEJ could be reproducibly detected in dried, human skulls and (b) clinical attachment levels could be reliably measured in human subjects. When using the Florida PASHA probe to detect the CEJ at 157 different sites in four dried, human skulls, there were no statistically significant (p > or = 0.15) differences in mean CEJ detection measurements for any of the three participating examiners; either when the CEJ was visualised or obscured. The mean differences between first and second replicates ranged from 0.00 to 0.08 mm. Intraclass correlation coefficients (ICCs) of repeated measures in both conditions ranged from 0.70 to 0.83 for tactile CEJ detection (CEJ obscured), and from 0.95 to 0.96 for visual detection (CEJ visible). In human studies, the Florida PASHA probe was used by two examiners to determine clinical attachment levels (CAL) at 660 sites in 5 human subjects undergoing supportive periodontal therapy. Intra-examiner agreement of replicate measurements recorded by the probe, as measured by calculating ICCs, ranged from 0.79 to 0.85 for the 2 examiners, respectively. A statistically significant inter-examiner difference in mean CAL measurements when using the Florida PASHA probe was found (p<0.001). Notwithstanding this difference, inter-examiner agreement was good, with an ICC of 0.83. These data suggest that the Florida PASHA probe can reproducibly detect the CEJ and is proposed as a tool for measuring CAL in humans.