The purpose of this study was to determine whether a pulse analyzer was useful 1) to characterize the variables of pulse wave of diabetes mellitus group, compared with those of healthy subjects, and 2) to determine Sasang Constitution in diabetes mellitus group and healthy subjects. 1. The sum of pulse pressure (energy) and the ratio of systolic period area (As%), called pulse pressure-related variables, were higher in diabetes mellitus group than healthy group, while the height of dicrotic wave (h5) and the ratio of height of dicrotic wave to height of percussion wave (h5/h1), correlated with arterial compliance, were lower in diabetes mellitus group than healthy group. 2. Taeumin of diabetes mellitus group showed higher pressure-related variables than that of healthy group. 3. Soumin of diabetes mellitus group had shorter the time to dicrotic wave (t5) than that of healthy group. 4. Soyangin of diabetes mellitus group showed higher heart rates and lower values in pulse wave time-related variables, including time to dicrotic wave(t5), time to incisura (t4), total time minus time to incisura (t-t4), total time (t), width of percussion wave (w), and the ratio of width of percussion wave to total time (w/t), than that of healthy group. 5. Contact pressure (CP), sum of pulse pressure (energy), height of pre-incisura (h2), height of incisura (h4), width of percussion wave (w), time to incisura (t4), time to percussion wave (t1), variance of total time (Vt), variance of height of percussion wave(Vp) and the ratio of height of incisura to height of percussion wave (h4/h1) were used to develop the rules of Sasang Constitution Classification with about seventy five percents accuracy. These suggested that the pulse analyzer was useful to evaluate the risk degree of diabetes mellitus and to determine Sasang Constitution among either diabetes mellitus group or healthy group.