Peres used Salecker and Wigner's prescription for a quantum clock in a theoretical study of time-of-flight determination of the velocity of a free particle. In the present paper, the quantum clock is applied to a closely related problem, determination of the average time spent inside a one-dimensional potential barrier V(z)Θ(z)Θ(d-z) by initially free “incident” electrons of energy E. For an opaque rectangular barrier straightforward application of the approach leads to a “clocked” result that differs by orders of magnitude from the result postulated by Büttiker for the average “intrinsic” dwell time τ D(0,d;E). It is shown that this difference can be eliminated by appropriate choice of initial state for the ensemble of identical clocks and by applying to their average behaviour when coupled to tunneling particles the calibration determined for the corresponding ensemble of freely running clocks. The difference that persists for more transparent barriers is attributed, following Peres, to perturbation of the clock and/or particle dynamics during the measurement process. It is most serious in the limit d→0 where the ratio of clocked to intrinsic times peaks at a value of about 1.6. Because of the nonlinear relation between “actual” time (i.e. the parameter t in the Schrödinger equation) and uncalibrated clock time it does not seem possible to decompose the mean dwell time for an opaque barrier as “measured” by the quantum clock approach considered here into individual components associated with transmitted and reflected particles. This is consistent with the point of view that the “tunneling time”, which refers only to the transmitted particles, is not a meaningful concept within conventional interpretations of quantum mechanics.