The utility of the Supplementary Data System (SDS) compiled by the Bureau of Labor Statistics (BLS) in conducting surveillance of occupational skin disease was evaluated by examining 14,703 workers' compensation cases reported to the SDS for the year 1981. Combined with state employment denominators obtained from the BLS Employment and Earnings Program, rates of illness (cases of dermatitis/10,000 employed) calculated for eight major industrial divisions varied significantly according to the criteria used for reporting cases. Despite quantitative variations in the rate of skin disease that depended on specific reporting criteria, the relative ranking of the major industrial divisions remained unchanged, with highest rates of skin disease consistently found in three major industry divisions: agriculture, manufacturing, and construction. This ranking of major industry divisions by rate of dermatitis corresponded extremely well with rankings generated from the 1981 Annual Survey (Spearman rank correlation = .98, p less than .01). At the two-digit level of the Standard Industrial Classification, the rankings based on the SDS had a 77% rank correlation with those from the Annual Survey. Two-digit SIC codes identified from the top 10 in both sets of rankings included crop and livestock production from the agricultural division and leather products, food products, rubber and plastic products from the manufacturing division.