The first objective was to investigate the transdermal iontophoresis of interferon beta 1b (IFN); the second was to determine whether the addition of 10 Arg residues at the N-terminus, creating a highly charged poly-Arg analogue (Arg
10
-IFN), increased delivery. Cumulative permeation of IFN and Arg
10
-IFN after iontophoresis at 0.5 mA/cm
2
for 8 h was 6.97 ± 4.82 and 9.55 ± 1.63 ng/cm
2
, respectively – i.e. >1000-fold less than that of ribonuclease A, cytochrome
c
and human basic fibroblast growth factor. Co-iontophoresis of acetaminophen showed that, in contrast to lysozyme, neither IFN nor Arg
10
-IFN interacted with skin to decrease convective solvent flow. Furthermore, there was no statistically significant difference between (i) iontophoretic delivery of IFN across intact or laser porated skin and (ii) passive or iontophoretic delivery of IFN across laser porated skin. Chromatographic characterisation supported the hypothesis that IFN was bound strongly to albumin. The formation of a ~ 86 kDa complex with albumin was probably responsible for the poor cutaneous delivery of IFN/Arg
10
-IFN despite the use of iontophoresis and/or laser microporation. Biopharmaceuticals might interact with specific proteins during iontophoretic transport and so decrease their (per)cutaneous delivery without affecting electroosmotic solvent flow, which is usually considered as a reliable marker to report on permeant binding during electrotransport across the skin.