When regression function belongs to Besov space, wavelet estimations are investigated by Chesneau and Shirazi (2014), Kou and Liu (2016, 2017). However, in many practical applications, one does not know whether the regression function is smooth or not. It makes sense to discuss consistency of wavelet estimator. This paper considers the mean L-p (1 <= p <= infinity) consistency of wavelet estimator for multivariate regression functions based on biased data.