In order to realize the echo online extraction for diffuse reflection laser ranging when echo photons are blocked by a large amount of noise photons, an echo online extraction method based on time correlation is proposed for diffuse reflection laser ranging system. Based on the characteristics of time correlation for diffuse reflection laser ranging data, the extraction problem is simplified into n points searching problem. Next, the correctness of the extraction based on time correlation is verified. And the rules to select a proper n are given further through the trade-off between search cost and extraction accuracy. Then, the evaluation criteria for short-time linearity and adjacent similarity are given in detail. Finally, Genetic algorithm is adopted to realize parallel search for buffer data and online data, decreasing the search time as well as increasing the extraction accuracy of ranging data. The echo extraction for Yunnan Station′s experimental data show that the extraction method proposed has strong ability to extract echo signal from high noise background quickly, which provides a feasible way to realize the signal online extraction for diffuse reflection laser ranging.