An historical perspective on hazard surveillance performed by NIOSH from 1970 to 1983 was presented. NIOSH has conducted two major national surveys as part of its hazard surveillance program: the National Occupational Hazard Survey (NOHS) conducted between 1972 and 1974 and the National Occupational Exposure Survey (NOES) between 1981 and 1983. NOHS was a survey conducted on a probability sample of 4,636 facilities excluding agricultural, mining, and government facilities distributed among 67 standard metropolitan statistical areas. All possible exposures to specific chemicals, trade name products, and physical agents were recorded as potential exposures during walkthrough surveys of the facilities. Basic information about each facility's business and occupational health policies was also collected. The data were extrapolated to the national workforce as a whole, enabling estimates to be made of the total number of workers exposed to a particular hazard, the proportion who used protective equipment, the percentage who were given periodic medical examinations, and similar outcomes. NOES was conducted because by 1979 data from NOHS was becoming progressively outdated. NOES was similar in its design and goals to NOHS. NOES was also designed to compile data together with similar data from NOHS in such a way as to allow trends in exposures to be determined. NOES surveyed a probability sample of 4,490 facilities in 98 geographic sampling units, including 1,830,330 employees in 523 different industries and 410 different occupations. More than 10,000 potential hazards and 100,000 trade name products used at the facilities were identified. Representative data obtained in the two surveys showed that the proportion of facilities having an established health unit at their site increased from 14% (NOHS survey) to 24% (NOES survey). The proportion requiring new employees to have a medical examination increased from 35% (NOHS) to 44% (NOES). A more than two fold increase in the proportion of facilities that perform environmental monitoring of work exposures occurred, from 10% at the time of NOHS to 28% at the time of NOES. The surveys have also revealed that many industries have improved their measures to control exposures. More than 99% of the facilities in the primary metal industry were found by the NOES to be controlling exposures to X-radiation versus 85% found by NOHS. The authors conclude that the trends revealed by the surveys may reflect factors such as the Occupational Safety and Health Act of 1970, employers' concerns about legal liability, employers' increased investments in human resources, and employees' increased awareness of workplace hazards.