Physical growth of 212 randomly selected Swedish urban children has been investigated from birth to sixteen years as part of a prospective longitudinal study of growth and development. Twenty body measurements have been taken at specified ages. In all, about 66,000 values of measurement have been recorded. At sixteen years 179 (84.4%) ofthe original children were still regularly being followed. Thirty-three (15.6%) of the 212 children left the study at various ages. A comparison between the children who left the study and those remaining did not show any significant differences in physical growth. All recorded data have been scrutinized in order to detect errors of measurement or administrative errors. Such values have been corrected or excluded. The editorial procedures also included the adjustment of each measurement to exact target age, interpolation of missing examinations and exclusion of children with an aberrant growth pattern. When comparing the present study with older Swedish investigations a secular trend was established, while there was good agreement with other contemporary Swedish investigations. Growth charts of distance values of various body measurements are presented. In the charts standard deviation lines (+/- 1, 2, 3 SD) and mean values are plotted. A logarithmic time scale (logarithmic conceptional age) has been used for both practical and theoretical reasons. When evaluating individual values in comparison with reference values standard deviations should be used rather than centiles. With this approach even grossly deviating values can be evaluated. Longitudinal follow-up of individual children and comparison of children of different ages will also be facilitated by this approach.