A high incidence of skin reactivity to antigens derived from unclassified mycobacteria (photochromogens, Runyon group I, scotochromogens, Runyon group II, and nonchromogens, Runyon group III) has been demonstrated throughout the United States, particularly in the southeast.1 It is presumed that at some time in childhood, many individuals are exposed to these agents, become infected and occasionally demonstrate clinical disease, and thereafter manifest skin reactivity. It is recognized that a wide pattern of antigenic cross reactivity occurs which makes determination of the inciting agent difficult and that cross reactions with standard tuberculin cause significant confusion in the detection of tuberculosis.2 The incidence of clinical disease in children associated with the acquisition of skin reactivity to unclassified mycobacterial antigens has only been revealed in scattered case reports without simultaneous surveys of the incidence of skin reactivity to appropriate antigens. The source of the infecting agents and the means by which