This paper addresses the problem of tracking random drifting parameter
s of a linear regression system. The asymptotic properties of several
estimation algorithms in the limit of slow drift are studied. The basi
c tool is the central limit theorem for a class of stochastic differen
ce equations established under weak conditions on disturbances and obs
ervations. The estimates of the rate of convergence obtained in the pa
per allow the asymptotically optimal algorithms to be developed.