Conducted in Iowa and North Carolina, the experiment included 809 families randomly selected and assigned to a control group or to 1 of 5 experimental treatments. Eligibility required a family income at the experiment's beginning of less than l 1/2 times the official poverty line. Of the 809 families, 729 remained in the program for the entire 3 years. Benefit formulas consisted of a basic benefit, a minimum level of income guaranteed to families with no other income; and an implicit tax rate, the rate at which the benefit was reduced as other income increased. Five different experimental treatments were used with basic benefit levels of from 50 to 100% of poverty level income and implicit tax rates ranging from 30 to 70%. Work and income responses to the experiment were examined separately for rural families whose income derived primarily from wages and for those whose main source of income was self-employed farming. On the basis of analyses which indicated significantly different response patterns by site and race, North Carolina whites and blacks and Iowa families (all whites) were analyzed separately. Effects of the experiment on attitudes and on nonwork behavior such as family stability, various forms of consumption, and school performance of children were also examined for the whole group. This report presents the results of this experiment. (NQ)