— In this paper we consider the problem of semiblind carrier frequency offset (CFO) estimation for OFDM systems. Specifically, we consider a synchronization scheme based on the insertion of pilot tones in the OFDM block. In this case, if the number of subcarriers is sufficiently large, the received signal vector can be modeled as a complex Gaussian random vector with a nonzero mean. Then, by exploiting the joint probability density function for complex Gaussian random vectors, semiblind maximum likelihood CFO estimators are derived. The performance of the proposed semiblind algorithms is assessed via computer simulations and compared with that of some blind CFO estimators previously proposed in the literature.