Numerous studies have focused attention on the influence of various biological and environmental factors on contact hypersensitivity. In order to evaluate the persistence and/or modification of allergic contact sensitivity to a number of common contact sensitizers, the same standardized patch tests were repeated on 174 subjects with contact sensitivity after a time lapse of 5 years (1987-1992). In 18.4% of the cases, 1 or more sensitivities were lost; 28.7% of the patients had a higher number of positive patch tests after 5 years, while the remaining 52.9% of the patients showed no change in the number of positive patch tests. In 88%, the positive allergens were unchanged, whereas in the remaining 12% of the subjects, they showed 1 or more variations. The association between the allergens most often positive was calculated for both the 1st and the 2nd patch test results. Moreover, to evaluate the frequency of an allergen's positivity, we studied the disappearance of old sensitivities and the appearance of new sensitivities by the McNemar test. Cobalt chloride was the only allergen with a significant frequency of new positivities over the period of observation (p < 0.01). Logistic regression analysis was performed to evaluate the possible influence of positive tests to other allergens, and of some clinical findings associated with contact dermatitis, on the sensitivity to cobalt chloride in 1987 and in 1992.