Chlorophyll a has traditionally been used as an indicator of phytoplankton concentration in natural waters. This parameter has also been widely used by phytoplankton modelers for the purpose of comparing model output with field data. Examination of 1974 field data from Saginaw Bay indicated that chlorophyll a concentrations were inconsistent with phytoplankton cell volume concentrations. Statistical procedures were employed to determine the significance and source of the observed variability in the ratio of chlorophyll a to cell volume concentrations. The procedures employed included t-tests, analysis of variance and polynomial regression. The hypothesis that the natural log of the ratio of chlorophyll a to cell volume was constant in time was tested on six different sampling station groupings representing various spatial divisions of Saginaw Bay. In all cases, this ratio was found to vary significantly at the 95% confidence level. The variability of the ratio decreased when only data from the outer portion of Saginaw Bay were used. Greater than 60% of the variability of the ratio can be removed by using a linear correlation of the natural log chlorophyll a cell volume ratio with the fraction of diatom cell volume.