Collection and storage of the clinical samples are crucial factors in the metabolomic workflows. However, with the expansion of metabolomics into the clinical domain and towards the large field studies in particular, the high sampling/storage standards practiced in the tightly controlled hospital environment cannot always be guaranteed. Thus, if the samples are exposed to suboptimal conditions and their integrity is compromised should they be discarded? Or such samples retain physiologically relevant information and can be of use? To explore the options we analyzed 117 urine samples that were collected under two different conditions. Part of the samples were collected within a clinical setting under optimal conditions, another part by patients at home and shipped to the hospital by mail. All samples were analyzed by liquid chromatography–mass spectrometry (LC–MS) and proton nuclear magnetic resonance (¹H NMR) spectroscopy. Multivariate modelling revealed clear differences between the two sampling conditions for both LC–MS and¹H NMR data sets. However, the differential metabolites appeared to be platform-specific, which clearly emphasizes the complementary nature of both techniques. The analysis of the samples that were exposed to suboptimal conditions revealed that age and body mass index remain as dominant traits of the metabolic profile, although their influence was stronger for LC–MS data. In conclusion, although it is important to ensure adequate sample collection and storage conditions, urine samples that do not fulfil these criteria still retain valuable physiological information and as such thus they could be of use for metabolomic studies when no alternative is available.