Hazard surveillance programs at NIOSH were discussed. The scope and problems associated with hazard surveillance were described. The historical background of NIOSH activities directed toward hazard surveillance was described. As a response to the Occupational Safety and Health Act of 1970 NIOSH conducted the National Occupational Hazard Survey (NOHS) in the early 1970s. The survey produced a list of nearly 9000 potential exposure agents and 86,000 commercial products containing approximately 72,000 different components. Data obtained in the NOHS were linked to the NIOSH Registry of Toxic Effects and Chemical Substances (RTECS) to produce a model capable of systematically identifying worker groups at high risk during the period 1975 to 1980. Because the RTECS contained fewer than 2000 of the 9000 potential exposure agents recorded during the NOHS, NIOSH supported research into a system designed to apply structure/activity relationships to compounds not listed in the RTECS system. This produced a system that can estimate potential toxicity for four endpoints: acute toxicity, carcinogenicity, teratogenicity, and mutagenicity. During the period 1980 to 1983 NIOSH conducted the National Occupational Exposure Survey (NOES). This yielded a list of 100,000 different commercial products of which approximately 50 percent had been identified as to their components by June 1988. This has resulted in more than 10,000 potential exposure agents being identified. The NOHS data have been reformatted as a job exposure matrix (JEM) which will enable it to be used in epidemiological studies. The JEM data have also been used with data from other national databases such as the National Health and Nutritional Examination Survey and the National Household Interview Survey. Typical uses of the NOHS and NOES data were described. Future NIOSH activities in the area of hazard surveillance were discussed.