Ability of human listeners to understand altered speech is argued as evidence of uniquely human processing abilities, but early auditory experience also may contribute to this capability. I tested the ability of Panzee, a language-trained chimpanzee (Pan troglodytes), reared and spoken to from infancy by humans, to recognize synthesized words. Training and testing was conducted with different sets of English words in natural, “harmonics-only” (resynthesized using only voiced components), or “noise-vocoded” (based on amplitude-modulated noise bands) forms, with Panzee choosing from “lexigram” symbols that represented words. In Experiment 1 performance was equivalent with words in natural and harmonics-only form. In Experiment 2 performance with noise-vocoded words was significantly higher than chance but lower than with natural words. Results suggest specialized processing mechanisms are not necessary to speech perception in the absence of traditional acoustic cues, with the more important factor for speech-processing abilities being early immersion in a speech-rich environment.