Marine ecosystems are increasingly threatened by overfishing and human-induced pressures that compromise stock assessments through species misidentification. In this study, we evaluate the efficacy of portable X-ray fluorescence (pXRF) as a rapid, non-destructive tool for species identification in fisheries monitoring, using marlin (Istiophoridae) as a model system. Anal fin spine samples from black, blue and striped marlin were collected at recreational tournaments and processed for elemental analysis. Multivariate analyses, including PERMANOVA, non-metric multidimensional scaling, and similarity percentage analysis, revealed significant interspecies differences in elemental composition. Notably, blue marlin exhibited distinct elemental profiles—potentially reflecting unique habitat use and trophic interactions—compared to black and striped marlin. Classification using canonical analysis of principal coordinates achieved correct classification rates of 68.0% for concentration data and 81.3% for raw spectral counts, while machine learning achieved classification rates of 51.5% ± 1.73 for concentration data and 62.6 ± 1.44% for raw spectral counts—each substantially exceeding the null hypothesis expectation of 33% accuracy for random guessing among three classes. These results underscore the ability of pXRF to discriminate species based on their elemental signatures and highlights the potential of pXRF as a cost-effective, field-deployable complement to traditional genetic methods in improving fisheries monitoring and conservation strategies.