Single dietary questions are used as a rapid method of monitoring diet. The aim of this investigation was to assess the performance of questions to measure population group intake compared to the mean of two 24-h recalls. Data from the Australian National Children's Nutrition and Physical Activity Survey 2007 was used (n = 4487). Children reported their intake on three questions relating to usual serves of fruit, vegetables and type of milk. Age, gender and body weight status were assessed as modifiers of the relationship between methods. There was a stepwise increase in fruit and vegetable intake (p < 0.001) measured by recall when grouped by response category of the short question. By recall, fruit consumption decreased with age (F = 12.92, p < 0.001) but this trend was not detectable from the short question (F = 2.31, p = 0.075). The difference in fruit intake between methods was greatest for obese children. Almost 85% of children who consumed whole milk by short question consumed mainly whole fat milk by recall, but agreement was lower for other milk types. Saturated fat and volume of milk was highest in whole milk consumers. Ease of administration suggests that short questions, at least for some aspects of diet, are a useful method to monitor population intakes for children.