The contribution of 222Rn from domestic water wells to indoor air was investigated in a study of 28 houses near Conifer, CO. Air concentrations determined by alpha-track detectors (ATDs) and continuous radon monitors were compared with the predictions of a single-cell model. In many of the houses, the water supply was shown to contribute significantly to levels of indoor 222Rn. The data from the ATD study were augmented with a continuous monitoring study of a house near Lyons, CO. The well water in that house has the highest known concentration of 222Rn in water yet reported (93 MBq m-3). The temporal pattern in the indoor 222Rn concentration corresponds to water-use records. In general, it is difficult to quantify the proportion of indoor radon attributable to water use. Several lines of evidence suggest that the single-cell model underestimates this proportion. Continuous- monitoring data, although useful, are impractical due to the cost of the equipment. We propose a protocol for 222Rn measurement based on three simultaneous integrating radon detectors that may help estimate the proportion of indoor 222Rn derived from the water supply.