A review of the control of carcinogenic exposures using the TLV approach presents a prospect of limited effectiveness. With asbestos, as with any carcinogen, no threshold is known below which no health effect may be manifest. At best, we have only limited dose-response information at levels much above those of practical concern. In the case of asbestos, current exposures can only be described crudely at any level of exposure, and health effects are only known for past high, but ill-defined, exposures. Limited information exists on the effects of synergistic interactions with other materials. The current U.S. TLV, based on data concerned with occurrence of asbestosis, has not been evaluated with regard to possible effectiveness in the prevention of asbestos cancer. Yet cancer is the heart of the asbestos-hazard problem. Finally, enforcement of the existing TLV, especially for asbestos has been limited, frequently absent, and often ineffective. Workers are exposed in many situations to levels much above the current standard. As discouraging as this picture may seem, a TLV can be useful for stimulating the development and application of engineering-control procedures. The application of these procedures, however, must be specified and mandated in future standards to lower worker exposures to the minimum commensurate with existing technology. As technology is developed that makes lower exposure levels possible in a large part of the industry, TLVs should be reduced to take advantage of that technology.