Selection of dietary assessment methodology is important in dietary intervention studies. Few studies have reported the relative merits of various assessment methods or the feasibility of electronic methods in pediatric populations. Parent-child dyads performed telephone recalls (no. = 32) and tape recorded dietary records (no. = 33). Traditional recalls were first conducted to familiarize parents and children with the telephone recall procedure, to introduce two-dimensional food models, and to instruct on the use of the tape recorder. Parents monitored and documented as unobtrusively as possible the child's intake on the day before the telephone recall and also on the day of the taped record. Children were called at random to reduce bias. Simple correlation coefficients (r) were calculated for nine nutrients and calories for both methods. For telephone recalls, values ranged from a low of r = .64 for saturated and polyunsaturated fat to r = .85 for cholesterol and r = .93 for iron. Tape-recorded data yielded r = .80 or above, except calories with r = .68 (p less than .001 for all values, 1-tail tests). Mean nutrient values were within expected ranges, e.g., 1,800 kcal +/- 500, with saturated fat about 14% of calories per day. Comparisons between parents and childrens reports of food frequencies and portion sizes revealed the best correlations for beverages, bread-cereals-crackers, meat-fish-poultry, and mixed dishes. We conclude that preadolescent children are able to provide dietary intake data using electronic methods in a manner that compares favorably with their parents' written records.