We have used microdialysis in the dermis for assessing penetration kinetics of salicylic acid (SA) in healthy volunteers (n = 18), following application on the volar aspect of the left forearm. Penetration was monitored at four locations: in normal (unmodified) skin and in skin with perturbed barrier function from (i) repeated tape stripping (ii) irritant dermatitis from 1 or 2% sodium lauryl sulphate (SLS) for 24 h and (iii) delipidization by acetone. The order of the treatments was randomized according to a latin square design. Epidermal barrier function and skin irritation were assessed in each location using evaporimetry and colorimetry. Transepidermal water loss (TEWL) values confirmed that both mild (acetone), moderate (1% SLS) and severe barrier damage (tape stripping and 2% SLS) had occurred. Microdialysis sampling with two parallel probes in the dermis was performed in each of the four treatment areas for every subject. SA (5% in ethanol) was applied in a chamber glued to the skin overlying the microdialysis probes and sampling was continued for 4 h. SA was detectable in all samples and measurable in all samples from penetration through perturbed skin. Comparing the SA penetration in barrier-perturbed skin with the penetration in unmodified skin in the same subject, the mean SA penetration increase was 2.2-fold in acetone-treated skin (P = 0.012), 46-fold in mild dermatitis and 146- and 157-fold in severe dermatitis and tape stripped skin, respectively (P < 0.001). The penetration of SA significantly correlated with the measurements of barrier perturbation by TEWL (P = 0.01) and erythema (P = 0.02) for each individual. Microdialysis sampling of SA penetration was more sensitive than non-invasive measuring techniques in detecting significant barrier perturbation in acetone-treated skin. A positive dose-response relationship for the percutaneous penetration of SA in response to increasing SLS pretreatment concentrations and thus the degree of irritant dermatitis was found. When analysing data by location on the forearm, a tendency towards an intraregional variation in the reactivity to barrier damage was found, with the most proximal location displaying higher reactivity scores than the most distal location in response to the same barrier perturbation procedures. The penetration of SA was not significantly different between locations. In conclusion, using microdialysis in the dermis to obtain real-time dermal pharmacokinetics in the target organ, this study demonstrates highly increased and differentiated cutaneous penetration of SA in barrier-perturbed skin. The measured drug penetration was demonstrated to correlate with non-invasive quantification of barrier damage.