Ethanol concentration quantitative analysis of ethanol-water solution can be realized by measuring the ratio of Raman characteristic peak heights. The content of ethanol can be determined by linear relation between relative intensity ratio and ethanol concentration. However, this analytical method only applies to the ethanol solution at low concentration. Concerning this issue, relative intensity of characteristic peak of ethanol (asymmetric stretching vibration of CH2 2 924.0 cm-1) and peak background of water (3 350 cm-1) at different ethanol concentration is experimentally measured by using a self-developed laser Raman ethanol content detection system. According to the relationship between relative ratio of characteristic peak heights and ethanol concentration, the nonlinear regression analysis methods are proposed to apply in the measurement of ethanol concentration in a wide range. Adjacent region average method is utilized to remove mutational random noise in Raman spectra of ethanol solution. Combined with multi-point interpolation processing, the baseline of Raman spectra can be calibrated. The influences of mutational random noise and the strong fluorescence background can be effectively eliminated with baseline correction and normalization methods. Polynomial and exponential mathematical models are adopted for nonlinear regression analyses by the relation between ratio of characteristic peak heights and concentration of ethanol solution. The analysis results show that the correlation coefficient of linear fitting and nonlinear fitting is about 0.991 and higher than 0.997 respectively. The linear analytical method can be effectively applied when ethanol concentration range is 15%～60%. The nonlinear analytical method has higher measurement accuracy in a wider ethanol concentration range of 3%～97%. Nonlinear mathematical model will provide theoretical basis for analysis of ethanol concentration, which can be applied in laser Raman ethanol content detection system to calculate the relatively accurate ethanol concentration of ethanol-water solution. Rapid, real-time and accurate quantitative analysis of wide concentration range ethanol solution, which has mutational random noise and strong fluorescence background interference, can be achieved by these analytical methods.