A modification of Somogyi's saccharogenic method for the determination of amylase activity has been presented for serum and urine. The starch substrate employed is buffered sufficiently to control the pH of sera and urines. A study was made of the effect of varying enzyme and starch concentrations, pH, chloride and buffer concentrations, and temperature. Also studied were various starches and other substrates, several methods for quantitating reducing substances formed, stability of samples, precision, and normal values. Identification of intermediate and end products was made by paperchromatographic technics. Above a limiting concentration of starch, representing saturation of enzyme with substrate, a linear reaction rate was observed for a period of time that was found to be directly related to the initial starch concentration. Kjeldahl's law of proportionality has thus been confirmed for the ra-amylase activity of serum and urine. Evidence is presented that deviation from a linear reaction rate occurs at about the time the starch and high molecular weight dextrins no longer saturate the enzyme. Following this point, hydrolysis of lower molecular-weight sugars gradually becomes the preponderant reaction.