The distortion product otoacoustic emission (DPOAE) corresponding to the frequency f2-f1 displays stereotyped, time-varying amplitude alterations during continuous primary tone stimulation. The origin of these alterations is unknown; however, evidence that efferent neurons contribute little to the changes has been presented (Kujawa et al., 1994a, 1995; Lowe and Robertson, 1995). The present investigation examines the hypothesis that these alterations in f2-f1 amplitude are a reflection of local, Ca(2+)-dependent mechanisms involving the outer hair cell (OHC) response to sustained stimulation. Experiments were performed using urethane-anesthetized guinea pigs with sectioned middle ear muscles. Intracochlear perfusion was employed to reversibly lower perilymph Ca2+ levels and to introduce antagonists and agonists of L-type Ca2+ channels. Manipulations that lowered available Ca2+ (zero Ca2+ artificial perilymph; zero Ca2+ with BAPTA) or that blocked its entry into the cell via L-type Ca2+ channels (nimodipine) reduced, prevented or reversed the perstimulatory changes in f2-f1 DPOAE amplitude. These perilymph manipulations also reduced the overall amplitude of this distortion component while perfusion of an L-type Ca2+ channel agonist (Bay K 8644) increased its amplitude. Mg2+ did not substitute for Ca2+, suggesting that these are not merely divalent cation effects. Results are consistent with the hypothesis that continuous stimulation-related changes in f2-f1 DPOAE amplitude are sensitive to perilymph Ca2+ levels and to the function of L-type Ca2+ channels. However, nimodipine also reduced the endocochlear potential (EP) and Bay K 8644 increased the EP. The sensitivity of both the perstimulatory changes in f2-f1 DPOAE amplitude and the EP to the latter drugs leaves their site(s) of action unresolved.