We examined how adsorption and desorption of gases from inlets and a cell could affect the accuracy of closed-cell FTIR measurements of carbon dioxide (CO2), carbon monoxide (CO), methane (CH4), nitric oxide (NO), nitrogen dioxide (NO2), methanol (CH3OH), acetic acid (CH3COOH), and ammonia (NH3). When standards were delivered to the cell through a stainless steel inlet, temporarily reduced transmission was observed for CH3OH and NH3. However, a halocarbon wax coated inlet (normally used on the system) had excellent transmission (comparable to room temperature Teflon) for both CH3OH and NH3, even at temperatures as low as 5¿C. Thus the wax is valuable for coating sampling system components that cannot be fashioned from Teflon. The instrument had a delayed response (~10--40 s) for NH3 only, which was attributed to passivation of the Pyrex multipass cell. To determine sampling artifacts that could arise from the complex sample matrix presented by smoke, the closed-cell FTIR system was intercompared with an open-path FTIR system (which is immune to sampling artifacts) in well-mixed smoke. A similar cell passivation delay for NH3 was the only artifact found in this test. Overall, the results suggest that ~10 s is sufficient to detect >80% of an NH3/CO ratio sampled by our fast-flow, closed-cell system. Longer sampling times or consecutive samples return better results. In field campaigns the closed-cell system sampling times were normally 10 to >100 s so NH3 was probably underestimated by 5--15%.