Abstract: The partial matched filter-fast fourier transform (PMF-FFT) algorithm can achieve fast parallel acquisition of spread spectrum signals. However, the scalloping loss introduced by its FFT part reduces signal acquisition performance. Based on an analysis of the scalloping loss improvement effects of traditional Hanning and Hamming windows, and addressing the insufficient improvement caused by their less-than-ideal maximum side-lobe peak and attenuation rate, this paper proposes a PMF-FFT signal acquisition algorithm improved by the Nuttall windows. Nuttall windows with different numbers of terms and orders were selected, and their amplitude-frequency characteristics were analyzed. Simulation experiments verified the scalloping loss improvement effect. Results show that the 4-term 5th-order Nuttall windows provides better improvement; it can reduce the scalloping loss introduced by the FFT part from −4.04 dB to −13.05 dB. Under a constant false alarm probability, the single-point detection probability of the signal increases from 20% to 98.16%, and the mean acquisition time decreases from 300 ms to 33.44 ms. This effectively improves the acquisition performance.