Urea has long been used to treat dry skin. In the present report, we compared two creams, identical with the exception that one contained both urea and sodium chloride and the other urea alone, in 22 patients with atopic dermatitis. Following a 2-week wash-out period, their clinically non-eczematous, rough or normal-appearing skin on the forearms was treated twice daily in a double-blind and randomised manner. We examined the treated areas by measuring transepidermal water loss, capacitance and electrical impedance. Our findings suggest that a moisturiser containing both urea and sodium chloride seems somewhat more effective than the same moisturiser without sodium chloride, at least concerning the ability to reverse impedance indices of atopic skin towards normal, an effect ascribed mainly to changes in hydration of the stratum corneum. However, the clinical significance of our impedance measurements is somewhat premature to decide.