The increasing measure of agricultural losses due to plant diseases caused by pathogens and pests is becoming a significant problem worldwide in the recent years. In order to produce enough food to support the population growth, early detection of plant diseases is imperative to reduce the crop spoiled during cultivation and harvest. Although many detection methods are available for diseases, they require either expensive instruments, cumbersome procedures or highly skilled operators. These disadvantages limit the applicability of these methods for on-field detection and confine them to the laboratory. Therefore, an early detection method that is different from the traditional practices is highly desired in the agricultural industry. Volatile organic compounds (VOCs) are largely produced by plants when infected by pathogens and / or infested by pests, and can be used as chemical markers for early detection of the onset of plant diseases. Therefore, electrochemical biosensor devices, which are capable of detecting plant diseases through measurement of VOCs, are proposed and established with the motives validated by the interviews conducted during a NSF funded I-Corps project. A biosensor for detection of 4-ethylguaiacol, a common VOC, was established using metal oxide (TiO2 and SnO2) nanoparticles. Another biosensor based on enzyme tyrosinase-immobilized on an electrode was successfully developed for detection of 4-ethylphenol. Methyl salicylate (MeSA), a VOC that plays important role in plant defense system, could be detected using alcohol oxidase / peroxidase-immobilized bi-enzyme biosensor after chemical hydrolysis. Another version of biosensor for MeSA detection was developed using a different bi-enzyme system involving salicylate hydroxylase and tyrosinase, which improved sensitivity (30.61 Acm-2M-1) and detection limit (13 nM). In addition, a tri-enzymatic biosensor consisting of an esterase in the electrolyte and salicylate hydroxylase / tyrosinase-immobilized screen-printed electrode were also developed for MeSA detection, yielding a sensitivity of 3.10 Acm-2M-1 and limit of detection of 750 nM. The platform for automatic VOC collection and temperature measurement for MeSA detection were developed using Arduino Uno and MOSFET. Finally, the enzymatic kinetic mechanisms were studied by initial rate measurements, and a mathematical model was developed to simulate the performance of the biosensor under various operating conditions.