Coffee-banana intercropping, widely practiced by smallholder farmers in South America and East Africa, is recognized for its potential to combine sustainability with resilience to climate change. This practice promotes crop diversification, but may also enhance water-use efficiency. However, its effectiveness may vary depending on the local conditions and agricultural practices. The lack of quantitative data on drought stress and the complexity of interactions within coffee-banana intercropping systems pose significant challenges in modelling and optimizing water use efficiency. This study aims to develop and refine innovative methods to assess drought stress in coffee-banana intercropping systems, with a focus on stable carbon isotope values (d¹³C), leaf temperature, and mid-infrared spectroscopy (MIRS). While stable carbon isotope analysis is a promising tool, its application may face challenges due to factors such as crop size, canopy heterogeneity, banana-coffee canopy overlapping, leaf age, orientation, or position (leaf morphological aspects), leading to variable competition for water and light. These factors affect the way sampling for stable carbon isotope and leaf temperature analysis should be conducted, in addition to physiological differences between coffee genotypes, agronomic practices, and complexities in data interpretation. Sampling and analytical protocols must be adapted to address these factors and their effects, while accounting for leaf morphology and microenvironmental parameters. Initially, we evaluated the influence of these factors on d¹³C variability in coffee leaf samples, in addition to their correlation with leaf temperature. Samples were collected from a 0.15 hectares experimental farm managed by the Agricultural Research Company of Minas Gerais (EPAMIG) in Brazil, an intercrop of Arabica coffee and Cavendish banana plants at 3.6 a distance apart. Coffee leaves were sampled using a metal puncher and leaf temperature was measured using