For the multisensor systems with unknown model parameters and noise variances, based on the system identification method and correlation method, the online information fusion estimators of model parameters and noise variances can be obtained. Substituting them into the optimal fused Kalman smoother weighted by scalars for components, a self-tuning fusion Kalman smoother weighted by scalars for components is presented.The proposed self-tuning Kalman smoother converges to the time-varying optimal fusion Kalman smoother in a realization, so that it has asymptotic optimality. It can be applied to self-tuning signal processing. A simulation example shows its effectiveness.