Abstract: Entropy minimization algorithm (EM) is a chemometrics method, which reconstructs pure spectra out from mixture spectra based on experimental data only, and no prior information needed. A rose essential oil was analyzed by GC/MS, two co-eluted mixture spectra were analyzed by EM, and unknown pure spectra were estimated and identified. The result shows that 8 pure spectra and their TIC concentration are estimated, and 4 of these pure spectra are identified by NIST library match in these two coeluted peaks. EM algorithm can analyze unknown components inside co-eluted peaks quickly, enhance the efficiency of nature product analyses, which is applicable in quick analyses of nature product, traditional Chinese medicine and food etc.