Objectives Increasing interest in using commercial nutrition apps to evaluate energy intake and diet quality warrants further investigation of how diet data are being analyzed by clinicians. Few studies have compared food composition databases from commercial apps with a validated nutrient database used by the scientific community. We investigated the reliability of MyFitnessPal (v19.4.0) and 2017 CalorieKing databases with the 2017 Nutrition Coordinating Center Nutrition Data System for Research (NDSR) database. Methods The 50 most consumed foods were identified from an urban weight loss study. A single investigator searched each database to document data on calories and nutrients (i.e., total carbohydrates, sugars, fiber, protein, and total and saturated fat). Intraclass correlation coefficient (ICC) analyses evaluated the reliability between each commercial database with the NDSR; an ICC ≥ 0.90 was considered excellent; 0.75 to < 0.90 as good; 0.50 to < 0.75 as moderate; and < 0.50 as poor. Sensitivity analyses determined whether reliability differed by most frequently consumed food groups. Results The 3 most frequently consumed food groups were Fruits (15 items), Vegetables (13 items) and Protein (9 items). There was excellent reliability between CalorieKing and NDSR for all diet data (ICC range = 0.90–1.00). Between MyFitnessPal and NDSR, calories and nutrients had excellent reliability (ICC range = 0.90–1.00), except for total fat (ICC = 0.89) and fiber (ICC = 0.67). Sensitivity analyses showed good to excellent reliability between both commercial databases with NDSR for Vegetables and Protein groups (ICC range = 0.86–1.00). However, MyFitnessPal and NCC had poor reliability in calories, total carbohydrate, and fiber within the Fruit group (ICC range = 0.33–0.43). Conclusions CalorieKing had stronger agreement with NDSR. The wide range of ICCs between MyFitnessPal and NDSR is attributed to poor reliability among foods within the Fruit group. Our findings illustrate variability in nutrient data across commercial databases that can impact use in clinical nutrition analysis and translation of evidence-based interventions into practice. Future studies are needed to compare other popular commercial mobile nutrition apps with NDSR and/or other validated nutrient databases. Funding Sources The first author is supported by the National Cancer Institute.