Purpose: Protocols for quantifying the surface roughness of implants are varied and dependent upon the roughness parameter produced by the particular measurement device. The objective of this study was to examine the accuracy and precision of typical roughness characterization instruments used in the dental implant industry. Materials and Methods: The average roughness (Ra) was measured using 2 common surface characterization instruments: an interferometer and a stylus profilometer. Titanium disks were prepared to represent 4 typical dental implant surfaces: machined, acid-etched, hydroxyapatite- coated, and titanium plasma-sprayed. Repeated measurements from multiple sites on each surface were undertaken to establish statistical inferences. Qualitative images of the surfaces were also acquired using a laser scanning confocal microscope. After surface measurements were conducted, the disks were diametrically cut and cross-sectional profiles were examined using a scanning electron microscope (SEM) as a comparative measure of surface topography. An analysis of variance was applied to isolate the effects of the measurement site, measurement sequence, surface treatment, and instrument type on Ra values. Results: The results indicated that surface treatment (P = .0001) and instrument (P = .0001) strongly influenced Ra data. By design, measurement site (diametrical: P = .9859; area: P = .9824) and measurement sequence (P = .9990) did not influence roughness. In the assessment of individual instrument accuracy, the interferometer was the most accurate in predicting SEM-based roughness (P = .6688) compared with the stylus (P = .0839). As a measure of aggregate precision over all measurements, the most repeatable instrument was the stylus (coefficient of variation [CV] = 0.108), followed by the interferometer (CV = 0.125) and SEM (CV = 0.273). Discussion: These results indicate dependencies in accuracy and precision related to the surface characterization technique. Conclusion: Instrument variability may obscure functional correlations between implant surface topography and osseointegration. INT J ORAL MAXILLOFACIAL IMPLANTS 2004;19:510–516