First principles, nonlocal density functional theory (DFT) calculations within the cluster approximation have been carried out in order to model the formation of Ti active sites in the pores of mesoporous silicas by the reaction of (eta(5)-C5H5)(2)TiCl2 with terminal surface silanols. In line with recent experimental observations (Maschmeyer, T.; Rey, F.; Sankar, G.; Thomas, J. M. Nature, 1995, 378, 159), the calculations predict that, after calcination, the dominant surface Ti species is the (=SiO)(3)TiOH complex. However, our results imply that there could be an appreciable concentration of the (=SiO)(2)Ti(OH)(2) complex present. The (=SiO)(2)Ti=O species is energetically unfavorable. Also in line with experiment, our calculations predict that the dominant surface Ti complex is formed via a (=SiO)(3)Ti(eta(5)-C5H5) intermediate. Calculated structures of this Ti intermediate are not, however, in agreement with the experimental EXAFS structure, and it is suggested that the cause of this difference is the use of an oversimplified model during the EXAFS fit. Computational studies of the hydrolysis and hydration of Ti centers provide a possible rationale for the uncertainties in previous studies concerning the nature and coordination of such sites in materials synthesized by different methods and studied under a variety of conditions.