Air-change rate is an important parameter influencing residential air quality. This article critically assesses the state of knowledge regarding residential air-change rates, emphasizing periods of normal occupancy. Cumulatively, about 40 prior studies have measured air-change rates in approximately 10,000 homes using tracer gases, including metabolic CO₂ . The central tendency of the air-change rates determined in these studies is reasonably described as lognormal with a geometric mean of 0.5 h^-1 and a geometric standard deviation of 2.0. However, the geometric means of individual studies vary, mainly within the range 0.2-1 h^-1 . Air-change rates also vary with time in residences. Factors influencing the air-change rate include weather (indoor-outdoor temperature difference and wind speed), the leakiness of the building envelope, and, when present, operation of mechanical ventilation systems. Occupancy-associated factors are also important, including window opening, induced exhaust from flued combustion, and use of heating and cooling systems. Empirical and methodological challenges remain to be effectively addressed. These include clarifying the time variation of air-change rates in residences during occupancy and understanding the influence of time-varying air-change rates on tracer-gas measurement techniques. Important opportunities are available to improve understanding of air-change rates and interzonal flows as factors affecting the source-to-exposure relationships for indoor air pollutants.