Two suggestions for sources of indoor 220Rn (thoron) have appeared in the literature: 1) building materials and outside air, and 2) soil beneath the house. Due to the difficulty of 220Rn measurement and limited data, both suggestions lack sufficient supporting evidence. We have investigated sources of indoor 220Rn in seven occupied houses in northern New Mexico, U.S. A two-filter system was used to measure indoor 220Rn levels continuously, and 220Rn progeny were measured with single filters and specialized alpha-track detectors. The amount of 220Rn entry from soil was curtailed by cutting off soil gas flow to the indoor air with subfloor depressurization mitigation systems. Four of the houses showed significant reductions in 220Rn with mitigation systems on. The average effect for these houses was to reduce indoor 220Rn levels by 70%. The other three houses had no clear reductions but in one of these houses, the mitigation system was not effective for stopping soil gas flow. Our results provide some of the most clear evidence to date supporting soil as an important source of indoor 220Rn.