The method of evaluating cold injury in woody plants by the use of triphenyl tetrazolium chloride has been refined to eliminate bias associated with visual differentiation between varying degrees of tetrazolium reduction and to predict tissue survival at a later date. An advantage of the method described here is that a small amount of tissue (50-100 mg) is required; this, therefore, allows for hardiness determinations at precise locations on the plant. The high correlation between cold injury and triphenyl tetrazolium chloride reduction may be due to cofactor and substrate limitations rather than inactivation of dehydrogenases.