A controversy over cancer risk has undermined trust in the scientific basis of regulation. Drinking water and pesticide programs are in chaos because "potential cancer risk" cannot produce practical standards. The risk controversy involves a dispute over the most fundamental scientific standards, namely what is and what is not a positive result. Here it is shown that the lowest effective dose (LOEL) of well-studied carcinogens is a firm and reproducible quantity that can serve as undisputed basis for safety standards. "Realistic risk assessment" is proposed, based on the immediately available LOEL, as a transitional measure to eliminate the controversy over assumed risks. A traditional regulatory safety margin, over which regulators have explicit authority, would produce traditional safety standards when based on the lowest effective dose. Superfund clean-up targets based on one in one-million potential cancer risk are equivalent to an approximately 10,000-fold safety margin over real risk. Realistic risk opens the way for a reevaluation of regulatory priorities based on the fact that a 5-fold safety margin for the real carcinogen arsenic is now in use and has proven safe.