Atopic dermatitis (AD) presents a large unmet need for treatments with better safety and efficacy. To facilitate development of topical therapeutics, we need an efficient model for assessing different formulations and concentrations. The "plaque model" has been successfully implemented in patients with psoriasis, another common inflammatory disease, to assess the efficacy of topical treatments. This model has not been validated for AD, which has higher placebo responses and less stable lesions than psoriasis. We aimed to assess changes in molecular signatures of intrapatient target lesions treated with topical therapeutics. We enrolled 30 patients with mild-to-moderate AD in a randomized, double-blind, intraindividual comparison of 3 approved agents applied blindly at the investigator site daily for 14days: pimecrolimus, betamethasone dipropionate, clobetasol propionate, and a vehicle/emollient control. Changes in total sign scores (TSSs), transepidermal water loss, and tissue biomarkers (determined by using RT-PCR and immunohistochemistry) were evaluated. TSSs showed improvements of 30%, 40%, 68%, and 76% at 2weeks with vehicle, pimecrolimus, betamethasone, and clobetasol, respectively, with parallel changes in transepidermal water loss (P<.05). Significant differences versus vehicle values were limited to steroids (P<.0001). Steroids (particularly clobetasol) restored epidermal hyperplasia and terminal differentiation versus minimal changes with vehicle or pimecrolimus (P<.001). Levels of cellular infiltrates and cytokines (IL-13, IL-22, and S100As) were similarly reduced only by steroids (P<.001). TSS improvement correlated with changes in hyperplasia, infiltrates, and differentiation markers. We detected significant clinical and tissue differences between agents, providing a novel approach to study the differential effects of topical formulations using a limited sample size.