We present MENA-150, a large-scale image dataset for the automated classification of traditional dishes from the Middle East and North Africa (MENA) region, showcasing 150 unique and rich local cuisines. These cuisines, with their unique geographic origins, are often underrepresented in existing datasets, highlighting a significant gap in food computing for regional and traditional dishes. Food classification poses a formidable challenge due to the vast number of categories, the high visual similarity among different dishes, and the lack of datasets for training deep learning models. Addressing these challenges, MENA-150 aims to bridge this gap by offering a dataset encompassing 150 fine-grained food categories with over 53,000 images, including 45,604 for training and 7,500 human-verified images for testing, thereby facilitating the evaluation of state-of-the-art models.