Metabolomics holds great potential for ecological studies, as a sensitive technique that generates extensive, complex datasets about the metabolic state of target species. A major challenge in achieving standardized results lies in sample collection and storage, particularly regarding biological material found in remote locations, far from research facilities. In such contexts, onsite sample preservation using liquid nitrogen or solvents can become unfeasible. Therefore, this study aimed to compare two preservation methods for plant samples, using bracken fern (Pteridium esculentum subsp. arachnoideum) as a model. To this end, twenty samples were collected from an bracken fern occurrence site within a Cerrado Brazilian Savanna conservation area, with ten samples assigned to each preservation method. Each sample consisted of a single frond from an individual plant. Two fronds were collected from each individual, one for each preservation method, each individual spaced at least 2 meters apart. The first preservation method was field preservation, with the immediate storage of samples in liquid nitrogen followed by storage at -20°C in an uninterrupted cold chain, and finally freeze-drying before grinding for extraction. The second method was herborization, by pressing samples between two sheets of filter paper using a botanical press for one month prior to grinding for extraction. The samples were extracted with methanol followed by a liquid-liquid extraction for chlorophyll clean up. The resulting extracts were transferred to vials for HPLC-DAD and UHPLC-MS/MS analysis. The chromatographic pattern of the herborized samples showed that the main phytochemical profile was conserved when compared to the field-preserved samples. While there were changes primarily in the peak intensities of major compounds, no significant new peaks appeared. Thus, the results indicate that herbarium-preserved samples are a promising alternative for plant sample collection in field studies, especially in remote locations.