Selaginella convoluta is a Brazilian endemic, desiccation tolerant plant with an effective drought-resistance mechanism. Herein, nuclear magnetic resonance (NMR)-based metabolite profiling was used to investigate this species in different hydration states: dried (harvested in the natural semiarid environment) and after rehydration. NMR is a suitable tool for this analysis because it allows the simultaneous detection of secondary and primary metabolites. To support NMR-based metabolite annotation, an in-house database was built by including all the experimental 1H NMR spectral data reported for the natural products previously identified in Selaginella spp. The analysis criteria were related to spectral similarity values based on the algorithm of pattern recognition between the in-house database and the total correlated spectroscopy (1H-1H TOCSY-NMR) experimental data. The in silico strategy with a similarity threshold of 50 %−100 % between the experimental and virtual spectra enabled the annotation of 86 metabolites. Among these, 49 metabolites showed a spectral similarity of 100 %. Notably, the isolation of the pyridine alkaloid anabasine from S. convoluta shoots and its structural assignment provided new chemical information for lycophytes. The multi-edaphoclimatic features of the Caatinga, an exclusively Brazilian semiarid biome, may contribute to an altered metabolite profile associated with desiccation. © 2021 Phytochemical Society of Europe