The percentage of cells dividing in a specific tissue of individual larvae can be estimated by analyzing DNA per cell by flow cytometry. An experimental test was carried out with cod (Gadus morhua) larvae, with brain as the target tissue, to validate this technique as an appropriate growth index for larval fish. Standard length (SL), myotome height, and %S-phase (% of cells in the S-phase of the cell-division cycle) variability were analyzed, with temperature (6 and 10 degrees C, food level (high- and no-food) and larval developmental stage (first feeding, pre-metamorphosis and post-metamorphosis) as independent factors. Cod larvae grew faster (in SL) and presented a higher %S-phase under high-food conditions. Larval SL increased with temperature in rearing and experimental ranks. However, there was a significant interaction between temperature and food in the %S-phase. There were no significant differences in the %S-phase between 6 and 10 degrees C at high-food levels. We suggest that this result is a consequence of temperature-dependency of the duration of the cell cycle. In the absence of food, larvae at 10 degrees C had a lower %S-phase than larvae at 6 degrees C, which may be related to increased metabolic costs with increasing temperature. Considering the effect of temperature, the mean % S-phase explained 74% of the variability in the estimated standard growth rate.