Global population forecasts dictate a rapid adoption of multifaceted approaches to fulfill increasing food requirements, ameliorate food dietary value and security using sustainable and economically feasible agricultural processes. Plant pathogens induce up to 25% losses in vegetable crops and their early detection would contribute to limit their spread and economic impact. As an alternative to time-consuming, destructive, and expensive diagnostic procedures, such as immunological assays and nucleic acid-based techniques, Raman spectroscopy (RS) is a nondestructive rapid technique that generates a chemical fingerprinting of a sample, at low operating costs. Here, we assessed the suitability of RS combined to chemometric analysis to monitor the infection of an important vegetable crop plant, tomato, by two dangerous and peculiarly different viral pathogens, Tomato yellow leaf curl Sardinia virus (TYLCSV) and Tomato spotted wilt virus (TSWV). Experimentally inoculated plants were monitored over 28 days for symptom occurrence and subjected to RS analysis, alongside with measuring the virus amount by quantitative real-time PCR. RS allowed to discriminate mock inoculated (healthy) from virus-infected specimens, reaching an accuracy of >70% after only 14 days after inoculation for TYLCSV and >85% only after 8 days for TSWV, demonstrating its suitability for early detection of virus infection. Importantly, RS also highlighted spectral differences induced by the two viruses, providing specific information on the infecting agent.