Measurements of the atmospheric mixing ratios of 10 nonmethane hydrocarbons (NMHC) and four halocarbons (methyl chloride, dichloromethane, trichloroethene, and tetrachloroethene) were conducted between January 1989 and July 1996 at Alert (Canadian Arctic, 82°27′N, 62°31′W). About 270 canister samples were analyzed covering the 7‐year period with an average frequency of about one sample every 9 days. The mixing ratios of these volatile organic compounds (VOC) exhibit considerable variability, which can partly be described by systematic seasonal dependencies. The highest mixing ratios were always observed during winter. During spring, the mixing ratios decrease for some compounds to values near the detection limit. The amplitudes of the seasonal variability, the time of the occurrence of the maxima, and the relative steepness of the temporal gradients show a systematic dependence on OH reactivity. The steepest relative decrease is less than 1% d−1 for methyl chloride, increasing to about 4% d−1 for highly reactive VOC. Similarly, the highest relative increase rates vary between 0.5% d−1 for VOC with low reactivity to 4% d−1 for reactive VOC. With the exception of ethyne, toluene, and methyl chloride the concentrations of all measured VOC decrease during the studied period, although this decrease is not always statistically significant. In general, the largest changes were found for the most reactive VOC, although the seemingly random overall variability observed for these compounds results in substantial uncertainties. For the less reactive VOC (ethane, benzene, and propane) the average relative annual decrease rate is in the range of a few percent per year. Dichloromethane and tetrachloroethene showed a decrease of 4 and 14% yr−1, respectively. The average decrease rate for the other alkanes is in the range of some 10% yr−1, indicating a substantial change of emission rates during this period. A likely explanation is a reduction in VOC emissions in the area of the former Soviet Union, most likely Siberia, as a consequence of the recent major economic changes in this region. The measurements were compared with the results of chemical transport models' simulations using the Emission Database for Global Atmospheric Research NMHC emission inventory. Although the model captures most of the main features of the shapes of the seasonal cycles of the NMHC, the results clearly show that model estimates are consistently too low compared to the observations. Most likely this is the consequence of an underestimate of the NMHC emission rates in the emission inventory.