The large temporal and spatial variability in carbon isotope fractionation of marine phytoplankton (%) is thought to reflect differences in environmental conditions. Meaningful interpretation of this variability requires an understanding of the processes responsible for phytoplankton isotope fractionation. While numerous factors have been suggested to potentially influence Epr recent theoretical and experimental evidence has emphasized the primary role of phytoplankton growth rate (mu) and CO2 concentration ([CO(2)aq]) in controlling epsilon(p),. Experimental examination of the relationship of epsilon(p) with mu and [CO(2)aq] in studies using different experimental approaches, however, has yielded inconsistent results. Here we directly compare new and previously published data on epsilon(p) as a function of CO2 concentration and growth rate for the marine diatom Phaeodactylum tricornutum. When grown under nitrogen-deficient conditions (nitrate-limited chemostat), epsilon(p) Of P, tricornutum decreases with increasing growth rate. In contrast, under N-replete conditions epsilon(p) values are considerably lower at comparable growth rates and CO2 concentrations and are largely insensitive to a 3-fold increase in growth rate due to increasing photon flux density. In both experimental approaches, epsilon(p) shows a relatively small CO2 sensitivity in the range of CO2 concentrations naturally occurring in the ocean (8 to 25 mu mol kg(-1)). Below ca 5 pmol CO2 kg(-1), a strong decline in epsilon(p) with decreasing [CO(2)aq] is observed. The apparent difference in epsilon(p) responses between nitrate-limited and light-controlled cultures of P, tricornutum suggests a principal difference in carbon acquisition for different growth-rate-limiting resources. A mechanistic explanation is proposed and potential implications for the interpretation of phytoplankton carbon isotope fractionation are discussed.