SUMMARY. 1. A model relating log chlorophyll a concentration to log epilimnetic total phosphorus (TP) concentration was re-examined based on: (a) comparative and temporal studies of four stratifying Wisconsin and other highly eutrophic temperate lakes; (b) comparative summer lake surveys from Iowa and Alberta.

2. Although P-limited, deeper lakes with long hydraulic residence times and low external and internal nutrient loading in summer had summer chlorophyll a yields below model predictions based on spring and summer epilimnetic TP concentrations.

3. For lakes with summer epilimnetic TP between 30 and 80 mg m−3, chlorophyll a concentrations exceeded model predictions based on summer TP. This relationship held even for Lake Delavan, Wisconsin, where the ratio of available N to P was unfavourably low during spring turnover, and where the trans-thermocline N:P flux ratio was sub-optimal for algal needs in early summer.

4. With increasing summer TP concentrations and/or increasing epilimnetic circulation depth (>5m), chlorophyll a concentrations fell below model predictions—independent of the potential for N-limitation. This plateauing in chlorophyll a response occurred at lower epilimnetic TP content (<c. 400 mg m−2) in lakes with elevated non-algal light extinction coefficients. Using Tailing's algorithm for the‘column compensation point’ (algal photosynthesis = algal respiration over diel cycle), light limitation best explains this fall-off in chlorophyll a yield.

5. The failure of the Dillon & Rigler (1974) spring TP v. summer chlorophyll a model for these Wisconsin lakes is unrelated to N-limitation. Instead, it reflects internal adjustment in take TP in response to stratification and seasonal external P loading.