To reduce the noise signal in the sampled data of the Electronic nose(enose), an improved adaptive method based on Kalman filtering algorithm is presented, which can adaptively adjust measurement error variance, while denoise the NonGaussian noise in an enose output. A Kalman filtering model for measurement data processing is introduced based on the signal characteristics in the enose measurement processing. The improved adaptive filtering method is discussed in detail and applied to enose signal processing. The results demonstrate that the proposed method can adaptively reduce the noise signal in the enose measurement data processing and improve the sensitivity of enose measurement.