In order to investigate N, P, K nutrient absorption and utilization efficiency in cassava-peanut intercropping system, five treatments on field experiment were conducted including mono-cropping cassava, mono-cropping peanut, one row cassava intercropping with two rows peanut(M1H2), two rows cassava intercropping with three rows peanut(M2H3) and two rows cassava intercropping with four rows peanut(M2H4). The results showed that: 1) The 3 intercropping systems showed intercropping advantage(IA) in yield, P and K absorption, N, P and K utilization. M1H2 and M2H4 showed significant IA in values. The yield equivalent ratio(Y) and value equivalent ratio(V) of the M2H4 were significantly higher than those of the M1H2 and M2H3. Y showed significantly positive correlation with P, K utilization efficiency(UP, UK), N absorption equivalent ratio(AN), P, K utilization efficiency equivalent ratio of cassava(UMP, UMK) and N absorption equivalent ratio of cassava(AMN) in the intercropping system. 2) Compare to mono-cropping cassava, cassava in 3 intercropping models showed IA in yield, value, P absorption, K absorption, and N, P, K utilization. Yield equivalent ratio of cassava(YM) showed significantly positive correlation with UP, UK, AN and UMP, UMK, AMN. 3) Compare to mono-cropping peanut, M1H2 and M2H3 showed significant intercropping disadvantages(ID) in the yield and value of peanut. There was ID in peanut P absorption and N, P, K utilization in M1H2, while there was IA in N, P absorption of the peanut in the M2H3. The yield equivalent ratio of peanut(YH) was significantly positively correlated with N, P, K utilization efficiency equivalent ratio of the peanut(UHN, UHP, UHK). In conclusion, M2H4 was the best model in 3 intercropping models.