Using temperature as an independent variable, we were able to explain some 43 and 49% of the variance in the specific absorption coefficient of phytoplankton (a*(lambda)) at wavelengths (lambda) 443 and 676 nm, respectively, for some 1187 samples collected over a broad range of oceanographic regimes. Through examination of ancillary data, we demonstrate that our results are consistent with the view that the size structure and taxonomic composition of phytoplanktonic communities are regulated by physical processes, for which temperature is often a suitable proxy. Results obtained from multiple-linear regression analysis showed that by using temperature and chlorophyll a concentration, both of which can be retrieved by remote sensing, a larger proportion of the residual variance in a*(lambda) could be explained than if chlorophyll a concentration were used alone. Furthermore, as a single independent variable, temperature could explain nearly an equal proportion of the variance of a* (lambda) as chlorophyll a. These results show clearly the potential of using satellite-derived temperature data to improve retrieval of phytoplankton biomass from remotely sensed data on ocean colour.