Membrane transport proteins (transporters) are critical players in the physiology of complex organisms. The range of metabolites a transporter translocates across a membrane (substrate spectra) determines the membrane selectivity and, consequently, the transporter\'s physiologic function. Nevertheless, determining the full substrate spectra of a transporter, and hence its role(s) in the organism, remains a daunting challenge. Many transporters lack an experimentally determined substrate, and many show a puzzling substrate promiscuity. Substrate promiscuity raises questions about which experimentally defined substrates are physiologic, which are not, and which are missing. Targeted approaches to determine transporter substrate spectra present limitations: Compound availability, lack of unexpected cofactors, and the researchers\' biases. To tackle these issues, transporters can be screened with complex native metabolites to generate new hypotheses on their function. Here, we describe an untargeted method to expand and compare the substrate spectra of sets of plant transporters. This method compares the substrate spectra of different transporters when expressed in Xenopus laevis oocytes and exposed to plant metabolic extracts. Comparing the transport-mediated accumulation of metabolites enables us to contextualize the relative transporter activity towards the same set of metabolites from undefined mixtures. Additionally, metabolite accumulation patterns across transporter sets allow us to annotate novel compounds. We validate the method with three well-characterized transporters, where we find unreported substrate spectra specificities and detect the transport of tens of metabolites at a time. We screen a set of fourteen transporters and find unreported activities for eight of them, some of which are confirmed from pure compounds. Finally, we propose to reinterpret the role of two plant transporters previously linked to developmental regulation (phytohormone transport) in light of their substrate promiscuity.