Data collected for the Edinburgh Lead Study have been used to investigate lead concentrations in children's naturally shed deciduous teeth. A within-child multiple-regression analysis has shown that the upper jaw has a higher concentration of lead than the lower, and that there is a gradient of decreasing concentration from the front to the back of the mouth. Even after the effects of jaw and tooth type have been allowed for, the concentration is still found to be negatively correlated with the weight of the tooth and with the age at which the tooth was shed. No statistically significant effects could be attributed to caries, fillings, or the incomplete resorption of roots. A single-valued index of tooth lead has been derived for each child, taking into account the fact that children gave different types of teeth.