Evaluation of adult literacy programs requires different criteria from those used to judge regular schooling. Indicators for evaluating adult voluntary programs with individualized curricula are recruitment; retention, including absenteeism; one-on-one tutoring with volunteers; training in a context; support services; the "quick-fix" syndrome; appropriate curriculum; and teachers who care. Common evaluation hazards are claiming much and providing evidence of little, selecting measures not logically related to the program, use of grade-equivalent scores, use of different instruments for pretesting and posttesting, and careless collection of data. Evaluations of programs in Philadelphia and Boston found that (1) low-literate adults often have highly developed coping skills and see little reason to upgrade their literacy level; (2) funding for illiterate adults is difficult to find because sources often expect yearly success rates, sometimes leading programs to accept only the better readers; and (3) quantitative tests tend to be geared to specific skills rather than relevant content. (Addresses and telephone numbers of seven resources are listed.) (CML)