Many of the current models for estimating the gross primary productivity (PPB) of terrestrial ecosystems depend on the efficiency of the use of radiation (EUR), the ratio of CO2 absorbed to the radiation absorbed. A methodology for estimating the EUR from satellite would be independent of the estimates of primary productivity from the climate data used so far to estimate the EUR. In this work we show that Mediterranean forests have very slight seasonal changes in vegetation indices, green biomass indicators and, conversely, large seasonal changes in CO2 absorption. The photochemical Reflection Index (PRI), calculated on the basis of reflectance obtained with the MODIS sensor, showed a high correlation with seasonal variability of the estimated EUR at ecosystem level with turbulent covariance techniques in two Mediterranean forests over five years. These results show us that it is possible to estimate the time variability of the EUR using the PRI and to improve estimates of carbon removals. This conceptual and technological advance thus avoids the need to make assumptions about variations in the EUR and the relationship with its climate controls.