Predictive value calculations were used to derive diagnostic guidelines for syphilis. Specificity of the VDRL (Venereal Disease Research Laboratory) and treponemal tests is high in healthy persons but less in elderly and ill persons. Sensitivity of the VDRL test is high in secondary and early latent syphilis but reduced in primary and late syphilis or in cerebrospinal fluid evaluations. Primary syphilis should be diagnosed by darkfield microscopy, with VDRL confirmation for atypical lesions. Screening of asymptomatic persons with the VDRL test, followed by treponemal test confirmation on positive sera, is recommended for all pregnant women, contacts of persons with infectious syphilis, and other high-risk groups. Quantitative VDRL assessment at 3, 6, and 12 months after treatment should be used to assess the adequacy of treatment for both late latent and early syphilis. Cerebrospinal fluid VDRL assessment and cell count should be restricted to seropositive persons with a high risk of neurosyphilis.