Relationships between standard plate count (SPC), laboratory pasteurization count (LPC), SPC after preliminary incubation at 12.8C for 18 hr (SPCPI), coliform count (CC), and psychrophile count (PsyC) on 758 raw farm bulk tank milk samples from a grade A supply have been determined. Linear correlation coefficients were: SPC vs. LPC, SPCPI, CC, and PsyC, r=0.39, 0.59, 0.48, and 0.56, respectively; LPC vs. SPCPI, CC, and PsyC, r=0.17, 0.24, and 0.18, respectively; SPCPI vs. CC, and PsyC, r=0.44 and 0.68, respectively; and CC vs. PsyC, r=0.55. Using an SPC standard of 100,000/ml as a reference, the following standards for the other four methods were found to be equal in terms of percentage of samples of the supply examined that exceeded the respective standards: SPCPI, 2,000,000; LPC, 3,000; CC, 1,000, and PsyC, 20,000 per ml. If unsatisfactory samples were considered to be those that exceeded the standard by one or more of the five methods, only about one-third of these were detected by each of these methods. This percentage was increased by use of paired combinations of the five methods. Of these combinations, the LPC-SPCPI combination detected the highest percentage (63.8) of the unsatisfactory samples. An SPCPI standard of 200,000/ml or a CC standard of 100/ml, as frequently suggested, would be approximately three times more severe in grading the supply examined than the SPC standard of 100,000/ml as specified in the 1965 recommendations of the Public Health Service.