Early diagnosis of fungal infections, which have seen an increase due to different environmental factors, is essential to an appropriate treatment of the plant by avoiding proliferation of the pathogen without excessive fungicide applications. In this work, we propose a microfluidic based approach to a multiplexed, point-of-need detection system capable of identifying infected grape cultivars. The system relies on the simultaneous detection of three plant hormones: salicylic, azelaic and jasmonic acids with a total assay time under 7 minutes, with LODs of 15 μM, 10 μM and 4.4 nM respectively. The three detection assays are based on optical transduction, with the detection of salicylic and azelaic acids using transmission measurements, while the detection of jasmonic acid is a fluorescence-based assay. The molecular recognition event for each metabolite is different: nanoparticle conjugation for salicylic acid, enzymatic reaction for azelaic acid and antibody-antigen recognition for jasmonic acid. In this work, two cultivars, Trincadeira and Carignan, presented infections with two fungal pathogens, Botrytis cinerea and Erysiphe necator. The grapes were tested using the microfluidic system alongside the benchmark techniques such as, high-performance liquid chromatography and enzyme-linked immunosorbent assay. The microfluidic system was not only capable of distinguishing infected from healthy samples, but also capable of distinguishing between different infection types.