Aiming at the slow calculation speed both of monochrome transmission based on Beer-Lambert Law and other transmittance models, the absorption features of oxygen A band was analyzed and algorithms of oxygen absorptivity in atmosphere was investigated in order to realize real-time and high-efficient passive ranging of infrared targets. The absorption spectrum and distribution characters for oxygen A band were analyzed. The passive ranging model based on random Malkums mode was built by introducing the random Malkums mode to the calculation of the band average absorptivity of oxygen molecules with the current atmospheric parameters, testing environment and HITRAN database. Then the passive ranging test platform was established for measuring seven different points ranging from 12 to 128 m. Experimental results show that the model is accuracy and the average error of oxygen A band between measured and calculated distances is 1.8% in current atmosphere. The proposed passive ranging algorithms is feasible and can satisfy the system requirements of rapid speed, online, concealing, higher precision, as well as strong anti-jamming.