The role of hair follicles in transdermal delivery remains difficult to elucidate due partly to animal model complications. This paper explores a novel technique employing two human skin membranes to differentiate shunt route delivery from bulk transepidermal input. The method monitors penetration through epidermal membranes and compares this with delivery through a sandwich of stratum corneum and epidermis, with the corneum forming a top membrane. As orifices of shunts occupy only 0.1% of the area, there is negligible chance that shunts in the membranes will superimpose. The top layer blocks shunts available in the bottom layer. If shunts are important, delivery through sandwiches will be much reduced compared with that through epidermis, allowing for increased double membrane thickness. Experiments with penetrants under passive, iontophoretic and electroporation conditions illustrated the value of the method. A Monte Carlo simulation suggested that any failure of membrane adherence would not affect conclusions drawn.