The further development of methods in the area of probation assistance is hindered by the unsystematic practices of probation assistants and by the lack of information about the effects of various procedures and programs. Efficiency studies are plagued by high subject drop-out rates, as well as by probation assistants' fear of further restrictions on their work and of scrutiny of their working methods. For successful research, it is essential that concrete hypotheses be posited before evaluation. Although non-institutional measures for probation assistance are considered superior to institutional efforts at resocialization, the relative effectiveness of various methods in different situations remains unclear. Evaluations must consider both the personality of the offender and the personality of the probation assistant. To do this, however, special personality tests for convicted offenders must be designed: such questionnaires must take into account probationers' usual lower class origins and propensity for falsification of data to project an image of social acceptability. Various tests must measure all areas in which treatment is expected to produce change. Another source of difficulty in evaluating effectiveness of measures is establishing success criteria. Overemphasis on recidivism may ignore progress made by probationers, changes in the crime patterns of gradually reforming offenders, and the significance of unreported crimes. Criteria selected must make possible determination of the suitability of particular measures for individual probationers at a given point and prediction of success chances. Further analysis is also needed on the attitude of probationers toward probation assistance and views of the public toward probation assistance and offenders. More psychological research on methods of probation assistance is required. Probation assistants themselves must actively encourage research and cooperate with research institutions to assure systematic development of resocialization methods. A 40-item bibliography is supplied.