The increasing popularity of green roof solutions stems from their numerous benefits, including improved building energy efficiency and urban environmental quality. However, the long-term success of these systems re-lies on the presence of healthy vegetation, which indeed is often subjected to urban stressors such as extreme temperatures and air pollutants. These stress factors frequently act over time and may not produce immediately visible effects, making continuous vegetation health monitoring essential. Reflectance Spectroscopy (RS) has emerged as a valuable solution for this purpose. RS methods provide versatile and scalable insights, enabling assessments from individual leaves to entire canopies and larger areas. RS data can be gathered and processed even by non-experts and automatically classified using simple spectral indices or advanced machine learning approaches. This overview explores the spectral characteristics of stressed vegetation and demonstrates how RS tools can detect these changes, detailing their configurations and associated challenges. Furthermore, spectral data classification techniques are presented, while current limitations and future research perspective are discussed. By addressing these aspects, this paper aims to serve as a roadmap for researchers and practitioners interested in leveraging RS techniques for monitoring the health of green roof vegetation.