In nutritional metabolomics a large inter-and intra-subject variability exists, and thus, it becomes important to limit the variance introduced by external factors. In a composite controlled study with full provision of all food for the standardized intervention, human urinary metabolite profiles were investigated for different factors, such as handling of urine collections, diet standardization, diet culture, cohabitation and gender. In study A, 8 healthy subjects (4 men; 4 women) collected 24-h urine, splitting each void into two specimens stored either at 4°C or at room temperature. In study B, 16 healthy subjects (7 men; 9 women) collected 24-h urine for three days while being on a standardized diet. Samples were analyzed by 1 H NMR and chemometrics. The NMR profiles indicated the presence of metabolites presumably originating from bacterial contamination in 3 out of 16 sample collections stored at room temperature. On the contrary, no changes in the NMR profiles due to contamination occurred in the 24-h urine samples stored at 4°C. The study also showed a trend towards a reduced inter-and intra-individual variation during 3 days of diet standardization. In study A, the urine metabolome showed a clear effect of diet culture and cohabitation, but these effects significantly attenuated after diet standardization (study B). Besides, gender-specific differences were found in both studies. Our results emphasize that best practice for any metabolomic study is a standardized, chilled sample collection procedure, and recommend that diet standardization is performed prior to dietary interventions in order to reduce intra-and intersubject variability.