The usual colorimetric method to determine a-acetolactate decarboxylase activity is long, is of poor accuracy (35%), and does not suit brewing fermentation conditions. By simple modification in the procedure, it was possible to improve the measurement to an accuracy of 11%, to allow more samples to be analyzed in a shorter period, and to reduce the cost per analysis. The recommended changes are easy to set up: the temperature of work should be decreased, the number of buffers should be reduced from two to one, and the chosen buffer should be maleate instead of 4-morpholineethanesulfonic acid. We also mention a way to measure both diacetyl and acetoin contents in the same set of measurement, based on absorbance readings of the reaction medium at two different times. These improvements are in accordance with brewing fermentation conditions.