A large number of factors are involved in the movement of ions and molecules across human epidermal membrane (HEM) under the influence of an electric field. These factors and their interplay need to be understood if our knowledge of iontophoretic transport of drugs across HEM is to reach a point where physical models and strategies may be employed for useful quantitative predictions. In a typical in vitro experiment, the fully hydrated HEM is positioned between aqueous compartments of a two-chamber diffusion cell. When a low electric field is applied across the HEM under these conditions, the transport enhancement of ions in the pre-existing pores of the stratum corneum is the result of, (a) the direct interaction of the electric field with the charge of the ion in question, and (b) convective solvent flow (electroosmosis); in the case where the permeant is non-ionic under these circumstances, transport enhancement is by convective solvent flow only. At moderate-to-high voltage iontophoresis (⩾around 1.0 V applied across a single HEM), in addition to the direct field effect and convective solvent flow in the pre-existing pores, there can generally be a significant (e.g. 10- to 100-fold enhancement) contribution to transport enhancement arising from new pore induction (electroporation). Much of the recent work in our laboratory has been devoted to defining and quantifying HEM electroporation, and an especially difficult aspect has been that of dealing with the large HEM membrane-to-membrane variabilities with regard to, (a) the extent of new pore induction, and (b) the characteristics of the newly induced pores. Recently we discovered that the extent of relevant (i.e. permeant accessible) pore induction may be correlated to the change in HEM electrical conductance (and quantifiable) if an appropriate matching background electrolyte can be selected having ion sizes comparable to that of the permeant. For example, employing tetraethylammonium (TEA) pivalate (PIV) for which the ion sizes are ≈3.5 Å, but not KCl (ion sizes ≈1.9 Å), as the background electrolyte for TEA (as the permeant) gave very good results; in this example, the sizable contribution of pore induction to iontophoresis was quantitatively factored out from the total iontophoretic enhancement. Experiments with a large number of HEM samples gave good agreement with the Nernst–Planck (N–P) predictions of the direct field effect when TEA-PIV was used as the background electrolyte for TEA transport, but large variations (up to 300%) between N–P predictions and experimental results were observed with KCl as the background electrolyte. Another area of recent effort has been HEM pore size determinations, both at low voltages (i.e. for pre-existing pores) and at voltages where the newly induced pores dominate HEM permeability. The sizes of pre-existing pores of HEM have been determined with the hindered diffusion theory (using experimental fluxes of several probe permeants of different known molecular sizes) to be generally in the range, 10–20 Å, by a number of investigators in our laboratory for a large number HEM samples. Deducing pore sizes of electric field induced pores under steady electroporation conditions has been a more challenging task. We succeeded recently in developing a novel method for ‘passively’ determining pore sizes (i.e. by passive diffusion with hindered diffusion theory) under steady electroporation conditions: by using low frequency (12.5 Hz) a.c. at 2–5 V. We have been able to sustain electroporation at a nearly constant state of electroporation long enough to carry out a set of ‘passive’ diffusion experiments with relatively good precision to obtain the sizes of the newly induced pores. Studies to date have revealed that the sizes of pores induced with 2–5 V are of the same order of magnitude as those of the pre-existing pores (i.e. 10–20 Å). Finally, another research question of interest has been that of pore charge (both the sign of the charge and the charge density) distribution of HEM. By conducting both anodal and cathodal electroosmosis experiments with a nonionic permeant, one can evaluate the predictivity of a model based on pores of a single size and single charge. Studies were conducted with HEM at low voltage (little or no electroporation) and at higher voltages (2–4 V). The results to date at neutral pH have shown that, at low voltage, while negative pore charges dominate, there are effectively some positive or positive and neutral pores. At high voltage where new pores dominate, there seem to be only negatively charged pores.