Nutrient profiling is a method that classifies foods based on their nutrient content and identifies foods that are high in micronutrients both across and within food groups. This study aimed to identify foods that are rich sources of the seven micronutrients (iron, zinc, calcium, thiamine, riboflavin, vitamin A, and vitamin B12) of public health concern for the Bangladeshi population.. This study developed a metric termed "naturally nutrient-rich score 7 (NNR7)" specifically for third-trimester pregnant women to identify nutrient-dense foods. Further, it computed the nutrient adequacy score (NAS) of the top NNR7-scored foods for seven micronutrients to assess the extent (percent) to which foods can meet pregnant women's recommended dietary allowances (RDA). A linear programming technique was then used to construct a nutrient-adequate model diet for third-trimester pregnant women using the top ten NNR7-scored foods. According to the NNR7, food groups such as leafy vegetables, fish, meat, poultry and eggs, and vegetables are the richest sources of the problem micronutrients. Mutton liver (916.7%), soybean (39.3%), lamb liver (2160%) and duck liver (50.0%) were found to fulfill the highest percentage of the RDA of vitamin A, zinc, vitamin B12, and iron, respectively. In the formulated nutrient-adequate diets for pregnant women, rice, potato, brown wheat flour, and soya oil were universal to all three diets and Bengal gram, orange, Ganges River sprat, and duck liver were the most common ones. The study findings highlight the need for the consumption of foods such as leafy vegetables, fish, meat, poultry, eggs, pulses and vegetables to increase the intake of problematic micronutrients. Planning a nutrient-adequate diet for pregnant women using linear programming can be an alternative approach to optimize and shape food choices to meet their nutritional requirements.