This paper reports a study of café-au-lait spots of a minimum diameter of 1 cm in 732 white schoolchildren. Three groups were identified, according to the number of café-au-lait spots on each child: (1) those with none (74%), (2) those with fewer than 5 (25%), and (3) those with at least 5 (5 children, 2 considered to be normal, and 3 siblings each presumed to have neurofibromatosis, one having died from leukaemia). Excluding the last group, the number of café-au-lait spots in the sample was not significantly related to age or sex. Some support is given for using the number of café-au-lait spots as an empirical threshold to diagnose neurofibromatosis.