Accurate dietary assessment is an essential foundation of research in nutritional epidemiology. Due to the weaknesses in current methodology, attention is turning to strategies that automate the dietary assessment process to improve accuracy and reduce the costs and burden to participants and researchers. ‘My Meal Mate’ (MMM) is a smartphone application designed to support weight loss. The present study aimed to validate the diet measures recorded on MMM against a reference measure of 24 h dietary recalls. A sample of fifty volunteers recorded their food and drink intake on MMM for 7 d. During this period, they were contacted twice at random to conduct 24 h telephone recalls. Daily totals for energy (kJ) and macronutrients recorded on MMM were compared against the corresponding day of recall using t tests for group means and Pearson's correlations. Bland–Altman analysis was used to assess the agreement between the methods. Energy (kJ) recorded on MMM correlated well with the recalls (day 1: r 0·77 (95 % CI 0·62, 0·86), day 2: r 0·85 (95 % CI 0·74, 0·91)) and had a small mean difference (day 1 (MMM − recall): − 68 kJ/d (95 % CI − 553, 418 kJ) ( − 16 kcal/d, 95 % CI − 127, 100 kcal); day 2 (MMM − recall): − 441 kJ/d (95 % CI − 854, − 29 kJ) ( − 105 kcal/d, 95 % CI − 204, − 7 kcal)). Bland–Altman analysis showed wide limits of agreement between the methods: − 3378 to 3243 kJ/d ( − 807 to 775 kcal/d) on day 1. At the individual level, the limits of agreement between MMM and the 24 h recall were wide; however, at the group level, MMM appears to have potential as a dietary assessment tool.