Detecting the activity of 210Pb in the human skull by counting its 46.5-keV gamma rays in vivo is a promising method to reconstruct one’s cumulative radon intake, based on which associated lung cancer risk can be evaluated. However, this technique is strongly challenged by the background radiation level, which can be largely categorized as room background and subject background. In this work, we quantitatively assess the performance of the phoswich detector in suppressing background radiation resulting from 40K ubiquitously present in human subjects under in vivo measurements using Monte Carlo simulations. We first determined the region of interest for 210Pb gamma-ray detection to be 31 to  61 keV and focused on the background level inside this region caused by two 40K decay processes. It is found that the 1.46-MeV gamma-ray–led background can be reduced by 40% by the phoswich detector operating in anticoincidence mode whereas the 1.31-MeV beta-particle–led background is almost unaffected. This observation is understood through the dependence of the anticoincidence efficiency on the incident gamma-ray energies. Our results suggest that the 1.31-MeV beta-particle–led background is much larger and harder to suppress than the 1.46-MeV gamma-ray–led background, and they call for more investigations in the background reduction techniques for 210Pb in vivo measurement.