The problem of the perilymph origin – influx of cerebrospinal fluid (CSF) versus ultrafiltration within the cochlea – cannot be solved by mere qualitative proofs of tracer passage through the cochlear aqueduct. In order to gain quantitative data on the possible perilymph sources, an experimental study was designed to follow the time course of dye concentrations in the cisternal CSF and in the perilymph after tracer injection into the CSF at the vertex. By comparing the resulting concentration peaks in both fluids, the mean peak of the perilymph tracer concentration was found to reach 36 % of the maximum CSF concentration only. It is concluded that the local perilymph production within the cochlea exceeds the influx of CSF by a ratio of about 2:1. A working hypothesis of the double perilymph origin is discus