Previous research showed that whitening the surface electromyogram (EMG) can improve EMG amplitude estimation. However, conventional linear-filter whitening fails at low EMG amplitude levels, perhaps due to additive background noise in the measured EMG signal. This paper describes am adaptive whitening technique, effective at low EMG levels, that cascades a non-adaptive whitening filter, an adaptive Wiener filter, and an adaptive gain correction. In experimental studies, subjects used real-time EMG amplitude estimates to track a uniform-density, band-limited random target. With a 0.25 Hz bandwidth target, either adaptive whitening or multi-channel processing reduced the tracking error to roughly half that of using force as the feedback signal. Increases in additive noise level, smoothing window length, and tracking bandwidth diminish the advantages of adaptive whitening.