Recent studies have focused on habitual intake without addressing meal-specific intakes. We aimed to identify meal-specific dietary patterns and their contribution to habitual dietary patterns. This cross-sectional study was conducted on 838 adults, both sexes who attended the health centres in Tehran. Dietary data were recorded by three 24-h dietary recalls (24hDR). Dietary patterns were identified by using principal component analysis on meal-specific and overall food intakes. Intraclass correlation (ICC) was used as a measurement of consistency across meals and days. Correlation analysis and linear regression (partial R 2) were used for meals contribution. Four habitual dietary patterns were derived from average dietary intake of 3-d 24hDR labelled as ‘Western’, ‘Healthy’, ‘Traditional’ and ‘Legume and broth’. Also, we identified two major dietary patterns on each meal level (factor 1 and 2 for breakfast, lunch, afternoon snack and dinner). The highest contribution of energy intake was observed in lunch (25·7 %), followed by dinner (20·81 %). Consistency of food groups was the highest across days (ICC tea = 0·58) and breakfasts (ICC tea = 0·60). Dinner had a strong correlation coefficient with the ‘Western’ habitual dietary pattern then followed by lunch. Similarly, dinner and lunch contributed the most (r and partial R 2) to the ‘Western’ habitual dietary pattern. Our results suggest that habitual dietary patterns to several extents are formed at meal levels, and dinner has a greater contribution to the habitual dietary patterns in Iranian people. This may help planning for local dietary guidelines according to the time of eating to promote public health.