An international survey of nitrogen dioxide (NO2) levels inside indoor ice skating facilities was conducted. One-week average NO2 concentrations were measured inside and outside of 332 ice rinks located in nine countries. Each rink manager also completed a questionnaire describing the building, the resurfacing machines, and their use patterns. The (arithmetic) mean NO2 level for all rinks in the study was 228 ppb, with a range of 1-2,680 ppb, based on a sample collected at breathing height and adjacent to the ice surface. The mean of the second indoor sample (collected at a spectator's area) was 221 ppb, with a range of 1-3,175 ppb. The ratio of the indoor to outdoor NO2 concentrations was above 1 for 95% of the rinks sampled, indicating the presence of an indoor NO2 source (mean indoor:outdoor ratio = 20). Estimates of short-term NO2 concentrations indicated that as many as 40% of the sampled rinks would have exceeded the World Health Organization 1-hour guideline value of 213 ppb NO2 for indoor air. Statistically significant associations were observed between NO2 levels and the type of fuel used to power the resurfacer, the absence of a catalytic converter on a resurfacer, and the use of an ice edger. There were also indications that decreased use of mechanical ventilation, increased number of resurfacing operations per day, and smaller rink volumes were associated with increased NO2 levels. In rinks where the main resurfacer was powered by propane, the NO2 concentrations were higher than in those with gasoline-powered resurfacers, while the latter had NO2 concentrations higher than in those using diesel. Rinks where the main resurfacer was electric had the lowest indoor NO2 concentrations, similar to the levels measured outdoors.