Infants with feeding problems requiring tube feedings may remain dependent on them for long periods. Experience with behavior therapy (defined as the extinction of specific problem behaviors at meal times and the fostering of adaptive eating behaviors) suggests that this may shorten the duration of tube feeding dependency. Objective: To compare the outcome of tube feeding dependency with behavior therapy and standard therapy. This is the first randomized trial of behavior therapy in this population.Methods: Subjects were tube feeding dependent infants aged 5 to 36 months and their primary feeder (32 infants in Group 1 and 31 in Group 2). Sample size was selected to detect improvement in successes with behavioral intervention (50%) vs. control group (10%). Exclusion criteria included recent aspiration, swallowing difficulties due to neurological or anatomical problem, and inability to attend regularly. Subjects were randomly assigned to one of two interventions: Intervention A (nutritional counseling + behavior therapy) or Intervention B (nutritional counseling). There were 7 weekly treatment and 4 follow-up visits. Total duration of involvement in the trial for each patient was 6 months. The primary outcome measure (success vs. failure of treatment) was the number of infants still dependent on tube feeding at the third follow-up visit in each group (i.e., 5 months after the start of the trial). The decision to discontinue tube feeding was made by an independent judge, without knowledge of intervention and using pre-defined criteria.Results: 63 infants were enrolled. The two treatment arms were similar with respect to infant age, birthweight, birth order, gender, development delay, severity of the feeding disorder, socio-economic status and parents' age and education. Because the study is not completed yet, it is inapproriate to break the randomization code at this time. However, preliminary results on the entire sample indicate that 10 out of 33 (30%) subjects at the third follow-up visit and 13 out of 30 (43%) at the fourth follow-up visit were successes. These success rates are in line with our predictions based on sample size and much higher than the rate expected without behavioral treatment. Final results by treatment intervention will be analyzed on completion of the study (March 1998). Conclusion: We suspect that behavior therapy will be found to be efficacious.