Recently, the use of adaptive filtering strategies such as the least mean square (LMS) and recursive least squares (RLS) methods have been employed to form MSE optimal detectors for CDMAwireless systems. LMS strategies suffer from poor convergence behavior, which is a function of the number of users in the system, and the inability to track fast fades. RLS techniques offer slightly better performance but don't exploit the inherent structure of the CDMA received sequence auto-correlation matrix to improve performance. This paper presents a new adaptive technique for MMSE in-terference suppression in CDMA systems motivated by lin-ear adaptive algorithms which are scalable in performance and complexity. Optimal performance is achieved when the scalable parameter, L, is matched to the number of users in the system. Extensive simulation results are presented for these techniques.