A laboratory experiment was performed to test whether differences in nutrient and energy demands between picophytoplankton and heterotrophic bacteria can explain the apparent inverse biomass relationship between these organisms in lakes along gradients of organic carbon and nutrients. Growth rates and final yield of cells were analyzed in crossed gradients of glucose and phosphate. Concentrations of phosphate (10, 25, and 60 microg P L(-1)) and glucose (0, 0.3, and 3 mg C L(-1)) were used in all possible combinations giving 9 different treatments. Heterotrophic bacteria had higher maximum growth rates in all treatments and became larger than picophytoplankton in many treatments. The variance in abundance of heterotrophic bacteria between treatments could almost completely be explained by the combined effects of glucose and P. In treatments where carbon limitation slowed down the growth of heterotrophic bacteria, picophytoplankton became abundant and these organisms showed a positive response to P in combination with a negative response to glucose. The negative effect of glucose on picophytoplankton is suggested to be indirect and caused by competition with bacteria that are favored by organic C. The results suggest that competition for phosphate between phytoplankton and bacteria is not size-dependent, that heterotrophic bacteria are superior competitors for P, and that organic C produced by picophytoplankton was of minor importance for heterotrophic bacteria.