The philosophy of the recently proposed "Levels of Laboratory Service" program, which will be so vital to the conduct of a successful outpatient tuberculosis treatment and control program, is presented. The hallmark of this program is the decentralization of the diagnostic/monitoring services as they involve laboratory participation. In the long run this could mean more efficient operation, more reliable reporting, and probably less work for the participating laboratories. The greater emphasis on smear examination (Level I) as a monitoring tool will mean fewer cultures, thereby lessening the load for those laboratories that once went through countless clinically requested exercises of repetitively proving by culture the existence of M. tuberculosis in a given patient. Doubtless, the bulk of the work will be conducted in Level II laboratories; but here, too, identification of the most easily defined pathogen, M. tuberculosis, will minimize the over-all workload for these investigators while decreasing their concern about mycobacteria other than tubercle bacilli. Expertise gained in frequent repetitions of a limited number of tests (niacin, nitrate reduction, and pH 7/68 degrees C catalase) will ensure reliable speciation of the clinically most important Mycobacterium. The work of Level III laboratories should eventually be reduced primarily to organisms other than M. tuberculosis, thereby ensuring that a number of highly competent reference institutions will not only attain proficiency in taxonomic aspects of mycobacteria, but will also reflect the regional picture of the changing patterns in mycobacterial pathogens of man. Participation of laboratories in proficiency testing programs will encourage top-level performance in all areas. Additionally, such testing programs will serve a teaching role; a laboratory need not feel "locked in" at a given service level, but may increase its proficiency and move up a step in terms of the service it provides. In contrast, no laboratory need feel compelled to increase its activities; if daily workloads limit the extent of their involvement with mycobacteria, these laboratories can be confident that other institutions are providing needed services. The success of the entire "Levels of Laboratory Service" program depends on the recognition by individual laboratories of their own workload limitation, the directed motivation of personnel, and the maintenance of a free and open pipeline of communication to laboratories at the next higher level of service.