Driven responses of cat auditory‐nerve fibers to long‐duration characteristic‐frequency (CF) tones could decrease substantially over time periods ranging from seconds to minutes. In extreme cases, the discharge rate could fall below the pre‐stimulation spontaneous rate (SR). Reductions in response were characterized by two processes, each of which followed a decaying exponential function. Long‐term adaptation affects the discharge rate in the first several seconds following stimulus onset. The average amount in high‐SR fibers was 42.5% for tones at 20–40 dB SL, and the mean time constant was 3.64 s. Long‐term adaptation increased significantly with sensation level (SL, or level above threshold), decreased with SR, and was not significantly correlated with CF or fiber response threshold. Time constants did not depend on CF, SR, or SL. Very‐long‐term adaptation refers to further, smaller reductions in the discharge rate that accumulate over a period of minutes. Fiber responses formed two groups. The larger group adapted with a mean time constant of 45.22 s for CF tones at 20–40 dB SL, and the smaller group did not adapt over very long terms.Considerable variability in amounts of long‐term and very‐long‐term adaptation was observed in similarly tuned fibers from the same ear. This suggests that long‐term effects do not arise from cochlear mechanics or middle ear muscle activity. No long‐term effects were observed in responses of fibers directly stimulated by high‐intensity electrical pulses presented at rates up to 500/s through a cochlear implant. This suggests that the effects do not arise from fundamental differences in spike‐generating properties of spiral ganglion cells. The data suggest that long‐term adaptation may occur either when neurotransmitter utilization at inner hair cell synapses exceeds the rate at which it is replenished from global stores or uptake mechanisms, or when metabolic resources influencing neurotransmitter release become depleted. The neural data are related to perceptual findings in human listeners, in which unusually large amounts of tone decay may be observed at high frequencies, and they indicate that the perceptual effects originate peripherally.