To further improve the anti-noise capability of laser methane remote sensors, on the basis of the property of “coherent detection, low 1/f noise” of wavelength modulation spectroscopy, we further use the ability of “multi-scale, multi-resolution analysis” of wavelet denoising to denoise the absorption signal. Herein, a methane remote sensor system was established, and the relevant parameters of wavelet denoising were optimized through simulations. Moreover, the effects of wavelet denoising based on empirical mode decomposition were discussed. Comparative experiments were conducted to verify the feasibility of the proposed technique in the sensor system for cases where the wavelet denoising algorithm was used and not used. For a methane gas sample with integral concentration level of 200×10 -6 m, measurement results showed that the signal-to-noise ratio was improved from 116.4 to 179.8 using the wavelet denoising algorithm to extract the second harmonic signal. The calibration experiments demonstrated that the amplitude of the extracted second harmonic signal was directly proportional to the gas concentration. For the cases where the wavelet denoising algorithm was not used and used, the goodness of fit was 0.990 and 0.996, respectively. Based on the Allan deviation results, the detection limit was 3.4×10 -6 m without using the wavelet denoising algorithm and 1.7×10 -6m using the wavelet denoising algorithm, with an improvement in detection accuracy by 1 times. The proposed laser methane remote sensing method based on wavelength modulation spectroscopy and wavelet denoising technology exhibited high signal-to-noise ratio, linearity, and stability. The proposed method can be popularized and used in existing methane remote sensor systems.