Surface water quality monitoring has traditionally relied on laboratory analysis of samples collected manually or using automated samplers. While laboratory analysis can provide data for a large variety of constituents including, but not limited to, metals, nutrients, and bacteria, this method of collection and analysis is both cumbersome and costly as well as insufficient if more frequent and spatially distributed data are needed to develop reliable estimates of pollutant loadings, fluxes, and trends. As the number of EPA 303d listings and approved TMDLs continues to grow, most notably concerning nutrients and bacteria, so does the demand for NPDES permit holders to collect and evaluate water quality data at a greater frequency than that afforded by traditional methods. Although various commercial water quality sensors have been developed over the last decade to meet detailed pollutant characterization needs, technology is not yet available for detection of many problematic pollutants.The objective of this study was to evaluate the potential for developing regression relationships between traditionally collected laboratory analysis data and water quality sensor data that would allow the approximation of continuous total phosphorus (TP) concentrations through the use of surrogate continuous water quality data. Correlations between TP concentrations and five surrogate parameters were analyzed during this study. The effect of logarithmic transformations of continuous and TP datasets on correlations was also examined. ANOVA analysis was performed on multi-parameter linear regression models developed. Finally predictive equations were developed using simple and multi-variate linear regression on surrogates identified in the correlation and ANOVA analyses. The results show that turbidity alone or paired with specific conductivity was the best surrogate continuous parameter for predicting TP.