The purpose of this retrospective study was to test whether the initial pattern of clinical presentation of shoe dermatitis could indicate the causative allergen(s) and to estimate the odds on foot dermatitis in patients with a positive patch test versus those with a negative patch-test result. Between 1990 and 2002, 8543 patients were patch tested with the standard series (and additional allergens, if appropriate). Of them, 1168 (14%) had been referred because of foot dermatitis and 474 of these patients (5.5% of the total group) presented a positive reaction to one or more substances related to shoes. We found that 6 standard allergens in the male group and 8 standard allergens in the female group were statistically significant for the shoe dermatitis group. The data showed a relationship between the distribution pattern of the foot lesions and most of the allergens. These results have clinical applications since the gender of the patients and the localization of the foot eruptions can, indeed, indicate what allergen is involved.