HPLC methods are now widely accepted as being the only accurate means of quantifying chlorophyll in aquatic systems. Very little is known about the comparability of HPLC with conventional in situ prompt fluorescence methods and with newer techniques involving the measurement of delayed fluorescence of phytoplankton (measure of living algal biomass) in aquatic systems. This paper investigates the use of HPLC for the calibration of in situ delayed fluorometric and Turner fluorometric methods and the correlation of these methods to one another when applied to chlorophyll measurements in waters from the Elbe estuary and the German Bight. The data shows that the correlations of HPLC to both methods were high r2 = 0.7-0.99 when all the samples taken were from the reaches of the river Elbe (605-725 km). The correlations of all methods were low at r2 = 0.45-0.52 for samples taken in the saline part of the Elbe plume and when the chlorophyll concentrations were low (0.2-9 μg · l-1) and with comparatively high chlorophyllide and chlorophyll c contents. Generally the correlations of HPLC to delayed fluorescence were better than the other correlations. This was probably due to the fact that delayed fluorescence measurements are not affected by the presence of other chlorophylls and their breakdown products whereas prompt fluorescence signals are. Delayed and prompt fluorescence methods can be calibrated accurately against HPLC values for chlorophyll in the natural samples and correlations remained good over several days. However, we suggest that it is advisable to check these as often as possible, particularly in the face of measurable changes of turbidity, salinity or spectral properties of the water. In this work the May and October slopes for the comparisons HPLC: delayed fluorescence in the limnic Elbe below Hamburg were similar (i.e. 223- 245). A significantly lower slope (160) was found for the 32 stations in the Elbe mouth/German Bight. This indicates that we were dealing with different water bodies and thus different algal populations and is backed up by the HPLC fingerprints of the samples. Our work shows conclusively that although a correlation may be good between absolute chromatographic methods and fluorometric/photometric methods, one cannot extrapolate a long-term conversion factor which holds for different sampling times or sites in any one system, not to mention between systems.