Cumulative and real-time performance of consumer-grade electronic radon monitors in a Canadian single-family home over three months.

Consumer electronic radon monitors (ERMs) are increasingly used for residential screening, yet long-term field evaluations under real-world conditions remain limited. This study conducted a 98day collocation of 15 models of ERMs (3 units each; 45 total) alongside a calibrated research-grade reference instrument (AlphaGuard PQ2000) and two alpha-track detectors (Radtrak 3 ; Radonova) in the lowest lived-in space of an occupied single-family home in Waterloo, Ontario, Canada, while concurrently recording environmental parameters, including PM2.5, CO2, temperature, and relative humidity. The average radon concentration reported by the reference instrument was 99 Bq/m 3 , with a range from 62 -141 Bq/m 3 . The cumulative and real-time ERM performance was assessed using a variety of metrics. Performance varied substantially across and within models. Eight out of the fifteen models had two or more units that achieved a relative percent error (RPE) within ±25%, with the top performers achieving an RPE of 3-5% after 13 weeks of deployment, while several other devices under-or overestimated, from negative 47% to positive 220%. Most models showed acceptable intra-model repeatability with a coefficient of variation lower than 15%. For real-time performance, top-performing ERMs achieved Root Mean Square Error of 5 -6 Bq/m 3 , and moderate (ρ = 0.70 -0.77), statistically significant Spearman-rank correlations on two-day non-overlapping averages. Overall, selected ERMs can approximate long-term cumulative average concentrations under real-world conditions, making them appropriate for long-term radon risk screening. However, the variability in ERM performance also highlights the need for regulatory oversight and independent testing to protect building occupants from a false sense of security or false alarms.