A Food Behavior Checklist (FBC) was developed, validated, and pilot-tested with homemakers in Virginia's Expanded Food and Nutrition Education Program (EFNEP). The FBC was designed to measure nutrition-related practices that are taught in EFNEP, but cannot be measured by the 24-hour recall. During development, 20 EFNEP paraprofessionals in Virginia and 20 randomly selected state EFNEP coordinators helped to identify items that are most important in evaluating the overall effectiveness of EFNEP, and which could be used to establish content, face, and construct validity of a behavior assessment instrument. The Food Behavior Checklist contained 30 items, and responses were recorded on a Likert scale with four response levels: 0 = "never or seldom", 1 = "sometimes",2 = "usually", and 3 = "almost always". Cronbach-alpha revealed a reliability coefficient of 0.86. During the field test phase of this study, the FBC was used before and after EFNEP intervention to collect data on the nutrition-related practices of 147 low-income homemakers in three rural counties and one urban area in Virginia. In this phase, the experimental group, which consisted of 79 homemakers, received computer-aided feedback on their dietary practices, via a Diagnostic Report, which was used as a teaching tool. Paraprofessionals thoroughly explained the content of the diagnostic report to the homemaker at program entry and program exit. The control group, which consisted of 68 homemakers, did not receive computer-aided feedback on their dietary intakes, in that no mention was made of the computer print out to them. Results indicated that this instrument (FBC) may be useful in evaluating the overall effectiveness of EFNEP nationwide and may be useful in other nutrition programs. No significant differences were observed in the nutrient intake or Food Behavior Checklist practice change scores between the group who received computer-aided feedback and the group who did not received computer-aided feedback. The author concluded that the dietary analysis contained in the computer generated Diagnostic Report needs to be simplified if it is to become an effective tool with EFNEP homemakers. More research is needed on the use of computerized diet analysis as a teaching tool with low-income homemakers. / Ph. D.