An unbiased evaluation of hospice effectiveness requires a randomized controlled trial (RCT), but the implementation of such a design is difficult. This paper describes the methods of the only RCT to date and reports on the validity and reliability of scales selected or developed to assess attainment of hospice goals. The patients involved are those in a VA hospital in Los Angeles who have cancer and a prognosis of about 6 months and who consent. They and their significant others are interviewed periodically until the patient’s death to measure such hospice variables as pain, depression, anxiety, satisfaction with care and the environment, activities of daily living, and bereavement effects. Many of the scales, including McGill Pain, CES-D Depression, and Satisfaction with Care scales, have been validated in other studies with similar samples. The authors’ test-retest reliability analyses show that these scales are reliable with advanced cancer patients and their significant others with coefficients ranging from 0.77 to 0.97.