There is perceived to be a problem of eutrophication in European marine coastal waters and hence a need to predict the response in terms of enhanced biomass of phytoplankton resulting from the input of anthropogenic nitrogen. This response was investigated indirectly by studying the relationship between nitrate and phytoplankton chlorophyll concentrations in waters of the Scottish west coast. Two-thirds of the 60 data sets analysed gave significant inverse regressions of chlorophyll on nitrate concentration. This result is explained by interpreting synoptic data as representing variation in time, with nitrate decreasing as a result of its assimilation by phytoplankton and conversion into chlorophyll-containing biomass Thus the absolute value of the slope of each significant regression estimates the yield (g) of chlorophyll from nitrate and, indeed, from any form of nitrogen assimilable by microalgae. The median value for q was 1.05 mg chl (mm01 N ) -' ; the range from 0.25 to 4.4 encompassed 95 % of values. Some, but not all, of the variation in q could be explained by error in individual estimates due to chemical-analytical and sampling errors or to inhomogeneities in each sampled phytoplankton population. The remaining variation in q included a seasonal trend, which might have resulted from changes in phytoplankton species composition, nutrient limitation status, or the balance between autotrophic and heterotrophic partitioning of nitrogen. It is suggested that an appropriate value of q can be used to predict the potential maximum increase in phytoplankton which would result from a given anthropogenic nitrogen discharge. The sensitivity of such predictions to error in determination of q is discussed, and the values obtained for q compared with observations in algal culture and mesocosms.