Phosphorus is a nutrient of particular interest in investigations of aquatic ecosystems due to its significance for water quality, biogeochemical cycles, and ecological stoichiometry. Two limitations to current methods hinder our ability to characterize freshwater phosphorus dynamics: inadequate detection limits and susceptibility to interference from silica. While the analytical constraints associated with detection limits are widely appreciated, the importance of silica interference is not. The magnitude of interference varies between analytical approaches used for colorimetric soluble reactive phosphorus (SRP) analysis because small variations in reagent chemistry, reaction time, temperature, and sample preservation affect the silica sensitivity. Variability in SRP methodology across individual laboratories makes it difficult to diagnose the severity of this issue in the literature. Here we show that standard methods (4500-P.F., APHA 1992) adopted for one automated method of phosphorus analysis results in significant overestimation of SRP in oligotrophic waters of moderate silica concentration (3.9 mu g P L-1 for 1 mg SiO2 L-1). We outline and apply modifications to this method to quantify low-level SRP (< 5 mu g P L-1) while eliminating silica interference for application in fresh waters, particularly those with a high background Si: P ratio. We adopt a more acidic color reagent, lower the reaction temperature, and increase the reaction time for a continuous flow autoanalyzer that is commonly employed in many freshwater ecology laboratories. We report a method detection limit (MDL) of 0.6 mu g P L-1 and show that the proposed method maintains sensitivity with negligible silica interference for investigators wishing to conduct low-level phosphorus analysis.