Accurate monitoring of tomato growth is essential for precision agriculture, contributing to improvements in crop yield and quality, while flexible wearable sensors enable in-situ monitoring of plant physiological information. However, their reliable application to fruits remains challenging due to the continuous deformation that occurs during fruit development. Here, we present a flexible fruit wearable system designed for in-situ and long-term tomato growth monitoring. Such a system enables real-time monitoring of in-situ reflectance spectra, along with complementary temperature and humidity data, throughout the tomato maturation process. Furthermore, machine learning algorithms are integrated to enable precise maturity classification, achieving a high accuracy (98.67%). To assess system performance, a 15-day thermal stress experiment was conducted, confirming its responsiveness under extreme environmental conditions. Additionally, a 36-day long-term monitoring experiment was conducted to assess continuous stability throughout fruit development, with the system maintaining functionality under large circumferential deformation. These findings underscore the system’s potential to advance precision crop management and productivity.