AbstractThe accuracy of several methods of age estimation based on developing teeth was tested on 63 individuals of known age between 0 and 5.4 years from the Spitalfields Collection. The following methods of age estimation were tested: atlas method of Schour and Massler; diagram of Gustafson and Koch; mineralization age of some deciduous and permanent teeth, Moorrees, Fanning and Hunt, a modified method of Moorreeset al.by Smith, and quantitative methods using deciduous tooth length and weight regression equations, Deutsch, Tam and Stack. Accuracy was measured as the difference between dental age from the developing teeth and actual chronological age available from parish records.Results show that the atlas and diagram methods are considerably more accurate for this population and age group than methods based on mineralization stages of formation. Accuracy for the atlas method was 0.11 (± 0.30) year and diagram method 0.10 (± 0.37) year. The accuracy for methods based on mineralization stages of formation were considerably less accurate: 0.52 (± 0.62) year for deciduous teeth, 0.57 (± 0.42) year for permanent teeth and 0.29 (± 0.39) year for the same method but modified specifically for prediction. Significant differences (P<0.01) between dental and chronological age were apparent for methods based on mineralization stages.Accuracy of the quantitative methods during the first year of postnatal growth was high, the most accurate being deciduous molar length at 0.02 (± 0.15) year.Factors influencing accuracy and the problems encountered with the methods tested are discussed. The Schour and Massler atlas remains the recommended method of dental age estimation for this age group for reasons of accuracy and