Objective: To evaluate the validity of self-reported weight and height measurements among residents of São Paulo, as well as the accuracy of these measurements for determining nutritional status, and to present calibration coefficients. Methods: A cross-sectional, population-based study was performed with a sample of 299 adolescents, adults and elderly of both genders, in São Paulo in 2008. Bland-Altman difference plot and intraclass correlation were used to determine agreement between measured and self-reported parameters. Sensitivity and specificity were assessed for overweight, and calibration coefficients were estimated for correction of weight, height and body mass index data. Results: The intraclass correlation was high between self-reported and measured parameters for weight (r > 0,94) and body mass index (BMI) (r > 0,85). The agreement between measured and self-reported weight, height and BMI was good. Sensibility was > 91% and specificity was > 83%. Conclusion: Self-reported weight measurements can substitute measured parameters in this population, in both genders and in the age groups studied. Self-reported height measurements should be used with caution. Calibration coefficients can be used to adjust self-reported measurements.