Traditional methods of estimating skeletal age at death have relied solely on the pubic symphyseal face or on this indicator combined with others in nonsystematic ways. A multifactorial method is presented that uses a principal components weighting of five indicators (pubic symphyseal face, auricular surface, radiographs of proximal femur, dental wear, and suture closure). This method has been tested by completely blind assessment of age in two samples from the Todd collection carefully screened for accuracy of stated age at death. Results show a marked superiority of the multifactorial method over any single indicator with respect to both bias and accuracy. This represents the first truly blind test of an age-at-death indicator or system, as the test populations were independent of the system(s) being tested, and the age, sex, and ethnogeographic origin of the individuals being assessed (as well as the compositions of the test samples with respect to these variables) were completely unknown until the tests were completed. Implications for paleodemography are discussed.