A theoretical framework was developed for analyzing Landry & Hassett (1982; Mar. Biol. 67: 283-288) type dilution assays under more general conditions, where phytoplankton growth might also be nutrient-limited. The theory, which accounts for growth on available internal and external pools as well as on nutrients .supplied through remineralization by grazers, was applied to a series of field experiments performed in nutrient-impoverished waters in the Oslofjord, Southern Norway, in summer 1988. Analysis of the experimental results indicates that the phytoplankton had intracellular stores of P sufficient for several hours of growth in all experiments, while internal stores of N were undetectable in 4 out of 6 cases. Dissolved N pools appeared to be of greater importance to the algae as a nutrient reservoir than were hssolved P pools. Significant nutrient supplies from remineralization were identified on 3 out of 6 occasions, compensating for around half of the estimated grazing losses in each case. The supplies of remineralized N and P were closely balanced with respect to the growth demands of the phytoplankton, indicating that the surplus of cellular P originated from other sources.