HYPOTHESIS: Intraoperative round window (RW) electrocochleography (ECoG) can help predict speech perception outcomes in adult cochlear implant (CI) recipients. BACKGROUND: Speech perception outcomes using CIs are highly variable. Recent data demonstrated that intraoperative ECoG could account for nearly half the variance in postoperative word scores. The present study seeks to update this correlation with a larger sample size and determine if addition of clinical variables improves the prediction. METHODS: Intraoperative RW ECoG was performed in adult subjects undergoing CI. Amplitudes of the ongoing response to tone bursts of multiple frequencies at 85 to 95 dB HL were summed to obtain the total response (ECoG-TR). ECoG-TR was correlated with postoperative speech perception scores. Multiple linear regression was used to combine clinical factors with the ECoG-TR. RESULTS: The ECoG-TR accounted for 40% of the variance in CNC word scores (n = 32). The preoperative pure tone average (PTA) was the only clinical factor with a significant correlation (r² = 20%). The ability to predict word scores using ECoG-TR and PTA, or after addition of age and duration of hearing loss, was not significantly different from using ECoG-TR alone. For 2 outliers, ECoG-TR predicted a better word score than obtained. CONCLUSIONS: The measurement of cochlear physiology before CI, reduced to a single variable, is a better predictor of postoperative speech perception than common clinical factors. Additional analysis of the outliers showed that waveform morphology can provide distinct information in individual cases.