Purpose. This paper examined urban household food demand patterns in Southwest, Nigeria using complement-substitution relationship. Methodology / approach. Through a multistage sampling technique, the study used cross sectional data of 445 households from two states in Southwest Nigeria which are representative of areas with a rapid rate of urbanisation. Quadratic Almost Ideal Demand System (QUAIDS) modeling framework was used to estimate the demand system for seven food groups considered. Results. The potential nutrition benefits of price and income changes in urban food groups was estimated, which explains the different quantity of food purchased by household as price changes. From the result, the root and tuber group accounted for the largest household food budget share (29.4 %) with the least share found in the legume group (3.3 %). Elasticity estimates revealed that the own-price effect was inelastic for all food groups. The uncompensated cross-price elasticities suggest both substitution and complementary association between groups. More substitution relationships were evident when households’ price changes are compensated. Originality / scientific novelty. This paper presents household’s responsiveness to food demanded through price effects with possible nutrition gap in urban diet. Practical value / implications. The complement and substitution relationship in household food price changes may be effective in changing urban consumer behavior towards healthier diets. Empirical knowledge of the complement and substitution effects would likely impact policy changes on household nutritional outcome. This is necessary as the scourge of malnutrition in increasing especially in urban areas.