We evaluated the role of the auditory–cognitive system in speech-in-noise perception in a group of older adults with hearing levels ranging from normal to moderate sensorineural hearing loss. In these 120 older adults (age 55 to 79), we used structural equation modeling to evaluate the strength of contributions from cognitive function (memory and attention), peripheral hearing status (audiometric thresholds and distortion product otoacoustic emissions), and neural processing (subcortical measures of pitch and response fidelity) to speech-in-noise perception (QuickSIN, Hearing in Noise Test, and Words-in-Noise [WIN] test). We also included a life experiences factor comprised of musical training because of its known long-term effects on speech-in-noise perception and memory and physical activity because of its effects on hippocampal volume and memory. We found that cognitive function and neural processing were the biggest contributors to variance in speech-in-noise perception, but life experiences also had an effect. Interestingly, the contribution of hearing thresholds was not significant. This finding is consistent with previous work demonstrating that the audiogram is not a good predictor of speech-in-noise perception. Implications for audiology practice are discussed.