Accurate and valid dietary data is the basis to investigate diet-disease relationships. Potential data discrepancies may be introduced when collecting and analysing data, despite rigorous quality assurance protocols. The aim of this study was to identify at-risk areas of dietary data in a food-based clinical trial. Source data verification was performed on a 10% random sample (n=38) of paper-based baseline diet history interview records in a registered clinical trial. All items listed in the source data underwent 100% manual verification based on the food input data from FoodWorks nutrient analysis software. Food item discrepancies were explored using food categories and summarised based on meals. The differences in identified discrepancies for energy and macronutrient output generated from FoodWorks software between previously entered data and re-entered data were compared. An overall discrepancy rate of 4.88% was identified. It was found that dinner intake data were more prone to discrepancy incidences than breakfast, lunch and snacks. Furthermore, assessing intake based on reported quantity and frequency may be more effective to correct discrepancies for quality improvement. Therefore, the dinner meal appeared to be an at risk area of dietary data. The method implemented in this study offers a systematic approach to evaluating dietary data in a research setting.