Determination of the drug susceptibility of Mycobacterium tuberculosis by conventional methods using an agar-based medium may take 3 weeks or more to complete. On the other hand, results on positive cultures are generally available in 4 to 7 days with the radiometric (BACTEC, Johnston Laboratories, Towson, Md.) procedure. One disadvantage to the latter is the requirement to determine the quantity of 14CO2 in each test vial on a daily basis from the day of inoculation. Growth index readings often must be made over weekends, adding to the work load of clinical laboratories during periods of reduced staff or necessitating compensatory pay or time. Susceptibility tests with streptomycin, isoniazid, ethambutol, and rifampin against 104 M. tuberculosis strains were performed by the submerged disk method, the recommended BACTEC method with daily growth index readings, and the radiometric procedure with readings delayed for 2 days after inoculation. Criteria for interpretation of "delayed" tests were established. Drug concentrations tested included some modifications of those available commercially. Overall agreement for the four drugs by the three methods was greater than 90%. We conclude that under our test conditions a schedule of inoculation of radiometric test vials on Friday with growth index readings commencing on Monday gives susceptibility results that correlate well with the daily BACTEC method and with a conventional 7H10 agar method.