The problem of separation of the shock wave and current sheet in magnetically driven shock tubes is examined. Experiments were performed in cold, unmagnetized hydrogen in a large, fast inverse pinch shock tube in which the simple ``skin‐depth criterion'' for obtaining separation was satisfied. Direct measurements of the shock by piezoelectric gauges, and the current sheet by magnetic probes, show that separation nevertheless did not occur. Although the shock wave was located at the very front of the current layer, the layer became much thicker at low pressures than predicted by the skin depth theory. Various mechanisms involved in the emergence of a shock from a current layer, and in the thickening of this layer, are discussed.