An instrument (TEMPO) has been developed to automate the most-probable-number (MPN) technique and reduce the effort required to estimate some bacterial populations. We compared the automated MPN technique with traditional microbiological plating methods and Petrifilm methods for estimating the total viable count of aerobic microorganisms (TVC), total coliforms (CC), and Escherichia coli populations (EC) on freshly processed broiler chicken carcasses (postchill whole carcass rinse [WCR] samples) and cumulative drip-line samples from a commercial broiler processing facility. Overall, 120 broiler carcasses, 36 prechill drip-line samples, and 40 postchill drip-line samples were collected over 5 days (representing five individual flocks) and analyzed by the automated MPN and direct agar plating and Petrifilm methods. The TVC correlation coefficient between the automated MPN and traditional methods was very high (0.972) for the prechill drip samples, which had mean log-transformed values of 3.09 and 3.02, respectively. The TVC correlation coefficient was lower (0.710) for the postchill WCR samples, which had lower mean log values of 1.53 and 1.31, respectively. Correlations between the methods for the prechill CC and EC samples were 0.812 and 0.880, respectively. The estimated number of total aerobes was generally greater than the total number of coliforms or E. coli recovered for all sample types (P < 2e⁻¹⁶). Significantly more bacteria were recovered from the prechill samples than from the postchill WCR or cumulative drip samples (P < 9.5e⁻¹² and P < 2e⁻¹⁶, respectively). When samples below the limit of detection were excluded, 92.1% of the total responses were within a single log difference between the traditional plating or Petrifilm methods and the automated MPN method.