[en] This report is concerned with improving estimates for the residence times of atmospheric trace constituents in various atmospheric reservoirs. Residence times are defined only for steady-state conditions; i.e., when the net growth rate vanishes. The most useful case of vanishing net growth rate is when the total growth rate is equal to the decay rate. It is demonstrated that the most important advance towards improving estimates of pollutant residence times is through proper choices of reservoirs. Chosen reservoirs should possess the following features: steady-state conditions, uniform mixing ratio throughout or throughout specified subreservoirs, and subreservoirs chosen in which removal rates can be treated as approximate constants. An example of a poorly mixed reservoir, the stratosphere, is discussed. In another example, it is suggested that commonly used reservoirs for atmospheric CO₂ have been chosen poorly and that a substantial portion of the anthropogenic CO₂ released during the past 50 years may still be mixing into the stratosphere. In another example, it is suggested that determination of the dry deposition velocity for accumulation-mode aerosol particles may not be so important as previously thought. To improve estimates for the atmospheric residence times of these particles, it is important to increase knowledge of what is called the ascension velocity