A double-blind, randomized, placebo-controlled study was conducted with 46 healthy female volunteers in order to test an anti-cellulite product containing retinol, caffeine and ruscogenine. An evaluation of different parameters related to cellulite appearance, i.e., the skin macrorelief, the dermal and hypodermal structures, the skin mechanical characteristics, and the cutaneous flowmetry was assessed using several non-invasive methods. This combination of different evaluation methods resulted in the demonstration of significant activity of the anti-cellulite product versus baseline and showed its superiority versus the placebo in skin macrorelief (decrease of the "orange peel" effect) and an increase in cutaneous microcirculation. By using a combination of methods, it was possible to detail the activity of an anti-cellulite product and to show superiority of the product in comparison with the placebo.