miology.Introduction: nutritional epidemiology seeks to understand the relationship between food consumption and the occurrence of diseases. During the COVID-19 pandemic, the Longitudinal Study of Adult Health (ELSA-Brasil) used a reduced Food Frequency Questionnaire (FFQ) to compare diet quality between cohort stages. Objective: to develop and validate an index created using a food quality score (FQS) using the FFQ of ELSA-Brasil. Methods: a cross-sectional study carried out between August 2020 and March 2021 with 4262 participants. For each frequency of consumption of healthy and unhealthy food markers, a score was given. Validity and reliability were tested using Pearson's correlation, principal component analysis (PCA) and Cronbach's alpha. Tests were performed to verify differences between individual characteristics and FQS. Results: the PCA indicated a total variability of 42.5 %. Women and individuals with higher age, physical activity and per capita income had a higher FQS, while those who reported higher consumption of alcohol and being smokers had a lower score. Cronbach's alpha was 0.47. Conclusion: the FQS showed a satisfactory validity and reliability and was associated with socioeconomic and lifestyle data. Therefore, it is a useful tool for comparing dietary information, thus contributing to nutritional epidemiology.