During a study of the carbohydrate distribution in grape canes, it became necessary to have a method for the estimation of small amounts of starch. A fairly rapid, accurate method yielding easily reproducible results was desired in order that the analysis of a great number of samples could be completed in a short time. An excellent survey of the literature dealing with analytical methods for starch is given by Walton (11). Of all the methods listed, one described by Thomas (9) for estimating starch in apple spurs seemed most suitable. The fundamental idea of Thomas's method is the direct determination of