This study proposes a new oscillometric-based blood pressure estimation algorithm. For credible blood pressure measurement, this study detects turning point of oscillation signal, obtained through digitaly filtered cuff signals which is of the subject collected from automatic blood pressure meter. Designed algorithm uses a characteristic ratios method from the turning point. The study compared the proposed algorithm blood pressure value and four kinds of blood pressure values of automatic blood pressure meter. Above results are compared by the value of mercury manometer that is manual blood pressure meter, in order to check performance of proposed algorithm. Regarding standard errors of SBP and DBP in automatic blood pressure meter obtained by comparing with manual blood pressure meter, wrist type blood pressure meter HEM-642 shows SBP ±8.76 and DBP ±6.47, forearm type blood pressure meter HEM-1000 shows SBP ±4.71 and DBP ±6.41, forearm type blood pressure meter FT-500R shows SBP ±7.29 and DBP ±5.35, and forearm type blood pressure meter Tensoval shows SBP ±5 and DBP ±4.59. The standard error of blood pressure value measured by manual blood pressure meter and proposable algorithm is ±3.35 in SBP and ±4.71 in DBP. The proposable algorithm suggested after examination is found to have lower standard error than four kinds of automatic blood pressure meters, and the proposable algorithm has credibility.