radon decay products in residential environments. It recommended that people measure indoor radon and consider taking action if the annual average concentration in their living areas exceeds 148 Bq m-3. EPA also developed programs in support of its recommendations for mitigation (Page 1993): public-information programs, a National Residential Radon Survey, Regional Radon Training Centers, the Radon Contractor Proficiency Program, the Radon Measurement Proficiency Program, Radon Reduction in New Construction, and support for the development of indoor-radon programs in individual states. As a result of those efforts, about 11 million of the approximate 100 million single family dwellings in the United States have been tested and about 300,000 (0.3%) mitigated in an effort to reduce indoor radon concentrations (CRCPD 1994). In addition, EPA estimates about 1.2 million new homes have been built with radon-resistant construction methods (A. Schmidt, personal communication), although the success of these methods is unknown.
It was recognized that water might also make a substantial contribution to and in some circumstances be the primary source of health risks associated with radon. In 1986, a revision to the Safe Drinking Water Act specifically required EPA to set a standard for 222Rn in drinking water (US Congress 1986). After litigation and a consent decree, EPA developed a criteria document that summarized the health effects of radon and its prevalence in drinking water (EPA 1991a). On the basis of the document and considerations of uncertainties in the analytic procedures for testing for radon in drinking water, a regulation was proposed in 1991 that established a maximum contaminant level (MCL) of 11,000 Bq m-3 (EPA 1991b). That MCL corresponded to an lifetime individual health risk of 10-4 posed largely by an increase in radon in indoor air.
During the period permitted for public response after the announcement of the proposed regulation, some groups supported reducing the MCL below 11,000 Bq m-3 because there is no known threshold for radiation-induced carcinogenesis. Others suggested raising the MCL because the increment in indoor-air radon from water radon at 11,000 Bq m-3 would be about 2% of the annual average residential radon concentration. There was also concern regarding the dosimetry model used to estimate the risk of stomach cancer associated with radon ingestion (Harley and Robbins 1994). As a result of those concerns, Congress intervened in 1992 and directed the administrator of EPA to prepare a multimedia risk assessment and cost estimates for compliance with regulations regarding radon in drinking water. The reanalysis resulted in EPA's revising its risk assessment for the ingestion of water containing radon. As a result, the ingestion risk and the inhalation risks (per unit of radon in drinking water) were estimated to be about equal (EPA 1994b). This document was reviewed by the Science Advisory Board (SAB) of EPA.
There was continuing concern about the estimates of stomach cancer resulting from radon ingestion. In addition, the SAB committee questioned the prudence of regulating a small increase in indoor radon from water without consid-