The widespread increase in the number of digital cameras mounted on flux towers provides an opportunity to better understand the relationship between the seasonality of canopy photosynthesis and canopy phenology. The challenge is due to fewer variations in rubber defoliation of rubber canopy. We examined the relationship between colour indices calculated from digital camera images and gross primary production (GPP) obtained from daily flux tower observations of carbon dioxide over two years in a rubber plantation and used these colour indices to model GPP. According to the results, (1) the strength of green (S-green), Hue and GPP exhibited clear seasonal patterns in the rubber plantation, and the relationship between the camera-based indices and GPP appeared to have distinct characteristics at different phenological stages; (2) the peak GPP from the flux tower measurements lagged behind the peak in the colour indices calculated from digital camera imagery, (3) GPP was strongly correlated with S-green derived from camera imagery in the rubber plantation, especially in the leaf expansion period; and (4) the GPP simulated by colour indices (S-green and Hue) was underestimated, and the fraction of absorbed photosynthetically active radiation (FPAR) was the best parameter for modelling GPP in normal years. Our results indicate that colour indices calculated from digital camera images can be used to model GPP in rubber plantations and to monitor biotic and abiotic stress events. Future research should measure the pigment contents of canopy leaves to precisely quantify the relationship between colour indices and GPP.