A reduced complexity quantization error correction method for lattice reduction aided (LRA) vector precoding is proposed. For LRA vector precoding,Babai's approximation procedure can generate quantization errors leading to performance loss. Instead of making a list to correct all possible errors as is done in the existing scheme, we propose a novel method in which only a subset of all possible errors are corrected. The size of the subset is determined by the probability distribution of the number of actual errors. Thus, the computation complexity of our correction procedure is reduced with little performance loss compared with the existing correction scheme.