A freshwater magnesium hydroxide coprecipitation method (MAGIC) has been developed to accurately and reproducibly determine low (nanomolar to subnanomolar) soluble reactive phosphorus (SRP) concentrations in freshwater. The method allows investigation of phosphorus distributions and cycling for systems in which SRP is below the detection limits of conventional methods. In natural waters, both inorganic and organic forms of P are coprecipitated; hence the method is essentially a preconcentration rather than a separation technique. Quantification of SRP on dissolved MAGIC precipitates follows a modified version of the standard molybdenum blue colorimetric method, using a spectrophotometer with ~0.1 milliabsorbance (mAbs) noise. Detection limit is 0.15 nM, improving on typical conventional colorimetric methods by a factor of ~50, with precision (RSD of triplicates) of ~10% at the 1 nM SRP level, 10% at ≤0.5 nM, and 4% to 7% at >1 nM. Considerable method development was necessary to eliminate or correct for multiple interferences, including a novel finding of potential interference by colored dissolved organic matter, and to optimize recovery, precision, and detection limit. The method was applied to filtered, frozen samples from western Lake Superior, showing that SRP concentrations are characterized by limited seasonal variability, largely uniform vertical distribution, and near‐bottom enrichment. Concentrations ranged from 0.4 to 10.9 nM SRP, representing ~10% of the total dissolved phosphorus pool. MAGIC is an easily employed analytical method appropriate for measurement of very low SRP in lakes and rivers.