Metabonomics is an important tool in understanding the toxicological or therapeutic effects of interventions by analysing metabolic profiles and interpreting complex multi-dimensional spectroscopic/spectrometric data using multivariate data analysis. The objectives of this study were to evaluate the metabolic changes following a short-term 5 day soya milk intervention, and to investigate factors that influence soy-phytoestrogen metabolism focused on Africans based in either UK or Nigeria. 1H-NMR metabonomics was applied to analyse urine samples collected at four phases I–IV (pre, days 3 and 5, and post) of the soy-intervention from African volunteers (n = 40 in total). Individual proton NMR spectra were visually and statistically assessed using multivariate analyses (MVA): principal component analysis (PCA) and (orthogonal-) partial-least square–discriminant analysis ((O-) PLS-DA). In addition, 22 endogenous metabolites were quantified using a Chenomx NMR suite. The results showed the levels of analysed endogenous metabolites (creatinine adjusted) present ranged from 4 µ m to 12 m m with large inter-subject variances in acetate, acetone, lactate and trimethylamine. The MVA results showed high inter-individuality and sampling variances based on PCA score plots, and demonstrated soy metabolism to be significantly influenced by location and gender by both PLS-DA and O-PLS-DA. Copyright © 2011 John Wiley & Sons, Ltd.