Three experiments were conducted using the California Consonant Test (CCT). Initially, performance-intensity functions were obtained for both normal hearers, and those with high frequency sensorineural hearing loss. Results demonstrated almost a linear function for both subject groups (approaching asymptote at 50 dB SL) as compared to the typical sigmoidal function obtained with conventional word recognition tests. Experiment II compared results of the CCT to those of the N.U. No. 6 lists in 60 subjects with high frequency hearing loss. Consistent with previous findings, relatively high word recognition scores were obtained for the NU-6 materials whereas the range of scores on the CCT approximated a normal distribution. Experiment III examined the internal consistency and split-half reliability of the CCT. Despite significant (p less than or equal to 0.05) correlation coefficients, the wide variability between half-list scores among individual listeners doses not support the use of half-list testing.