A few years ago, Professor A. M. Kligman posed the question ''Is dry skin dry?'' Since that time, our knowledge of dry skin has improved considerably, but it is still not clear whether or not the water content of dry skin is lower than that of normal skin. This is, however, a fundamental question that involves one of the most important problems in cosmetics. In the near infrared, water molecules show two clear absorption bands at 1450 and 1936 nm. The amplitudes are sufficiently high for them to be easily identifiable in the spectrum of the human skin in vivo. We have modified a commercial apparatus (Infra-Alyser 500; Bran and Luebbe) in order to obtain a system suited to spectroscopic analyses of all skin sites in vivo. Special optic fibers with a very low coefficient of attenuation and an integration sphere have been added to the original apparatus without modifying its capacities of function. In a first set of in vitro experiments, we examined the penetration of IR waves into the various skin layers. The results showed the role of the stratum corneum (SC) in the amplitude of the two water peaks and that the amplitude is proportional to the water content. The dry skin type known as winter xerosis is characterized by a marked alteration of the appearance and surface state of the SC. This led us to describe the clinical condition in terms of the following criteria: inflammation, roughness (smooth/rough), and presence of flakes and scales. We attributed a score of 0-4, and an overall score was then calculated. The water content of the skin was measured in the same area (external surface of the leg) in a number of subjects, using two biophysical methods-near-infrared spectroscopy and an electrical conductance method that provides only a very indirect measurement of water content. Correlations between the biophysical measurements and the clinical scores showed clearly that the skin judged as being the dryest gave the lowest peak amplitudes in the near infrared. This correlation was much better than that obtained using the clinical scores and the conductance method, particularly for very dry skin. A more thorough analysis of the correlations between the clinical scores and the NIR parameters provides further information on this phenomenon. From the previous population, three groups of 19 subjects with very dry skin were further studied. The subjects in each group applied a moisturizing preparation to one leg, the other leg serving as a control. The state of the skin was tested after four weeks of daily application by means of the near-infrared and conductance methods as well as by using the clinical scoring system. The results indicate that the water content of the skin and its conductance changed in the direction opposite to the macroscopic aspect. Near infrared spectroscopy emerged as the most sensitive method for evaluating the efficacy of the preparations. We present the overall results in terms of the clinical description of dry skin, together with the actual efficacy of the moisturizing preparation relative to the expert evaluation.