Compressive sensing and processing delivers high resolution data using reduced sampling rates and computational effort compared to Nyquist sensing and processing. Compressive processing, however, results to an increase in estimation error, which is particularly high when using a non-adaptive compressive sensing scheme. In this work, an adaptive compressive acquisition method is proposed for target tracking that utilizes information on target state available from the sequential tracking process. The proposed method reduces the peak ambiguity function sidelobe level compared to a nonadaptive method, thus improving tracking performance while adding little complexity to the compressive receiver.