Abstract: The regression analysis technique (REG) and difference technique were used to estimate endogenous P outputs and true P digestibility value with soybean meal (SBM) or wheat middling(WM) as only P source in diets for animals. Eight barrows, average initial body weight 20 kg, were allotted two groups randomly, fed four diets each group according to a 4×4 Latin Square design. Four cornstarch-based diets containing four levels of TP (0.15%, 0.20%, 0.30% and 0.35%) based on fed status each group were formulated from SBM or WM,respectively. Each experimental period comprised 8d with 4 d adaptation and 4 d collection feces. Results showed that: There were significant linear relationship (P＜0.05) between fecal P outputs and dietary inputs of P when both expressed as g/kg DM intake diet with TP level of diet from 0.15% to 0.35%, which suggests a stable amount of endogenous P lose in feces for pig. No significant differences (P＞0.05) of endogenous P existed determined by REG and difference technique with SBM and WM, implicated that the two assays are valid for estimating endogenous P outputs in pigs. The endogenous P output was 0.65~0.68 g/kg DM intake diets under tested TP levels and true P digestibility value of SBM and WM in growing pigs were 52.1% and 63.7%, respectively.