For the foreseeable future toxicologists will continue to rely on animal bioassays as the main predictive tools for safety evaluation. While their predictive value is good, there are serious limitations which must be recognized in applying the test results to assess human risk. These limitations are of especially great concern for chronic toxicity tests. During the past decade numerous protocols and guidelines for chronic toxicity and carcinogenesis tests have been developed by national and international organizations. These have been reviewed with two aspects in mind: the potential for standardization and the assessment of issues of greatest concern. Of the major design aspects, animal models, route of exposure, dose selection, pathology, and reporting requirements are the most controversial. While a degree of standardization of protocols has many advantages, complete standardization is not considered desirable as test objectives may differ, and overly rigid protocols would likely stifle development of improved methodology. One of the greatest dilemmas facing government health agencies is the extent of pathology to require in routine bioassays. A thorough microscopic examination is obviously desirable; however, the increased requirements for testing chemicals using the extensive pathology protocols now in existence has created a demand exceeding available scientific manpower.