This investigation was designed to study the correlation between behavioral audiograms and recordings of cortical potentials in experimentally deafened cats. Cortical thresholds were defined as the minimum intensity of tone pulses capable of eliciting strychnine potentials. Curves expressing threshold elevation with respect to normal animals were prepared for both the behavioral and cortical responses; these graphs were proportioned in accordance with the anatomical frequency scale and displayed above charts of cochlear histopathology. A mean value of 3.0‐db greater loss for the behavioral test was obtained when the differences between corresponding pairs of behavioral and cortical threshold elevations at all frequencies in all animals were averaged algebraically. The standard deviation of this collection of threshold differences was 17.4 db. Hypotheses advanced to explain the discrepancies include the effects of: (1) uncontrolled physiological variance in the cortical recording procedure; (2) the fact that the absolute cortical threshold exceeds the absolute behavioral threshold by 30 to 40 db; (3) the use of an abrupt onset for the cortical tone pulse. It is concluded that the cortical test is the most practical and descriptive electrophysiological indicator of threshold cochlear function presently available although it cannot be considered a precise substitute for behavioral threshold determinations.