Experiments were conducted to test the assumption, always made in models of the productivities of higher plant communities, that photosynthetic rates measured in steady light can be used to predict the rates in fluctuating light. The photosynthetic rates of whole cucumber plants (Cucumis sativus) were measured in light which alternated between high and low levels, at irradiances and frequencies comparable with the fluctuations found in nature (16—220 w/m2 of photosynthetically active radiation, 1012—103 sec at each level). The mean photosynthetic rate in alternating light was always within a few percent of the mean of the two photosynthetic rates in steady light at the two irradiances which were alternated. In a practical test, daily totals of carbon dioxide assimilated in natural fluctuating light caused by clouds were within a few percent of those calculated from the irradiances and the steady—state photosynthetic rate. In both experiments, the plants acted as near—perfect integrators of photosynthate, not of light. The conclusion is that steady—state photosynthetic rates may safely be used to compute the rates in fluctuating light, over the range of light conditions likely to be found in nature, provided that photosynthetic rates, and not irradiances, are integrated.