This paper focuses on the experimental and theoretical study of drug release from microemulsions structured as a dispersion of oil droplets in water or vice versa. In particular, drug release kinetics is studied by means of permeation experiments where a synthetic thin (149 mum) membrane separates the donor compartment, filled by the drug-loaded microemulsion, from the receiver one, filled by a drug-free aqueous-phase. Experimental trials evidences a peculiar release kinetics resembling that typical of permeation through a thick membrane where the permeated drug amount shows a not linear trend with time. This phenomenon is explained supposing the existence of an interaction between the drug and the surfactant micelles pervading the microemulsion aqueous-phase. This hypothesis is supported by independent experimental tests (critical micellar concentration measurements and release tests from a microemulsion containing small amounts of drug and surfactant) and by means of a mathematical model describing the whole permeation phenomenon. Nimesulide (anti-inflammatory action) is chosen as model drug for its industrial relevance while isopropyl myristate (oil-phase), benzyl alcohol (co-surfactant), Tween 80 (surfactant), compose our microemulsion (45.7% (w/w) water, 30.8% (w/w) surfactant, 11.75% (w/w) oil-phase and 11.75% (w/w) co-surfactant). The results of CMC measurements, release tests from low drug content microemulsion jointly with the good model data fitting ensure about the reliability of our hypothesis.