In Time Division-Synchronous Code Division Multiple Access (TD-SCDMA) system, the Steiner algorithm is used to perform channel estimation for the serving cell based upon the midamble in each timeslot, and joint detection (JD) is employed to eliminate intra-cell interference for both uplink and downlink. However, since single-cell JD (SJD) only deals with intra-cell interference and is not able to cancel out inter-cell interference, the users on cell edge typically achieve much worse link performance than those near the base station. Therefore, multi-cell JD (MJD) is proposed to further mitigate inter-cell interference. In this paper, the performance of multi-cell channel estimation (MCE) and multi-cell joint detection (MJD) techniques for TD-SCDMA evolved downlink are investigated and analyzed. It is shown that minimum mean-square-error (MMSE) multi-cell channel estimator and MMSE multi-cell equalizer can reduce downlink block error ratio (BLER) for the users both on the cell center and on the cell edge.