The diagnosis fibromyalgia (FS) requires the existence of tender points, routinely identified by clinical examination. We evaluated the interrater reliability of digital (thumb) examination for tender points by comparison with dolorimeter examination, a procedure considered to measure accurately muscle tenderness. Subjects were 15 patients with varying rheumatological diagnoses and anatomically widespread pain. In a physician blinded procedure, 2 rheumatologists determined the tender point count by digital examination at 18 points, and the tender point threshold by dolorimeter at 12 points. A pain threshold of 4 kg/1.77 cm2 or less defined the presence of tender points under both methods. Results indicate (1) classification as FS vs other diagnosis using pain complaint and digital examination for tender points, was moderately reliable (kappa = 0.74, p < 0.005); (2) interrater agreement about presence/absence of tenderness at individual points was not significantly lowered by digital examination (kappa = 0.51, p < 0.0001) relative to dolorimetry (kappa = 0.62, p < 0.0001); however, (3) analyses on the 12 anatomical points that were common to both methods indicated that digital examination resulted in significantly more anatomical points being considered tender relative to dolorimetry. Our findings indicate that digital and dolorimeter measures are equally reliable, but have poor concurrent validity for defining tender points in FS. Implications of these findings for the classification of fibromyalgia are discussed.