The basilar membrane response was calculated by cascading two‐port ABCD matrices that are formulated to include effects due to scalae tapering, perilymph viscosity, and helicotrema. Previously, Koshigoe et al. Acoust. Soc. Am. 74, 456–492 (1983)] showed that perilymph viscosity has an important effect on cochlear input impedance (Zc) in the frequency region below 500 Hz. It is shown that Zc is affected more by scala tapering than by perilymph viscosity, and it is affected very little by the helicotrema. It is also shown that low‐frequency standing wave (LFSW)'s seen in Zc of several models [J. W. Mathews, Ph.D. thesis, Washington University, Missouri (1980)] are strongly dependent on the CMF used. LFSW's present in nonlinear time domain solutions are likely to manifest themselves across the entire frequency spectrum and as a result mislead one's conclusions regarding basilar membrane mechanics. Previous models have used a straight CMF of the form K0 exp (− ax) resulting in LFSW's. Liberman's CMF leads to a reduction of LFSW's and when Greenwood's drooped CMF [J. Acoust. Soc. Am. 33, 1344–1356 (1961)] is used LFSW's are not present.