This paper measures the volatility of manufacturing employment growth across Canadian regions for the period 1976-97. It also attempts to relate the structural characteristics of regional economies to their levels of employment volatility. In particular, the analysis focuses on testing whether diversity, growth, plant size and export intensity are empirically related to volatility levels. Using cross-sectional analysis, the paper shows regions that are more stable tend to be more diverse, have lower than average growth rates, larger plant sizes and higher export intensity. These relationships are stronger for regions that have low manufacturing employment than for those where manufacturing employment is larger. Different results emerge when changes in volatility between the first and the last half of the period (1976-87 and 1988-97) are examined. Although increased diversity and export intensity are associated with decreased volatility in larger regions, these variables have the opposite effect on smaller regions. We therefore conclude that the admonitions about diversity and the benefits of increasing trade liberalization are not felt equally. They may be beneficial in terms of reducing volatility in larger regions, but they have not done so in smaller regions. The very nature of the small regions makes their adjustment to structural change much more difficult. Finally, the paper shows that increases in export intensity affects the underlying structures of regional economies. Regions with high export intensity and increases in export intensity are more specialized and becoming more specialized, respectively. Higher export intensity and increases in export intensity are also associated with larger and increasing plant sizes, respectively. The movement to freer trade between Canada and other countries then is associated with structural shifts in the Canadian economy that have offsetting effects on volatility. Increased trade is associated with larger plant sizes, which tend to dampen volatility, and decreases in diversity, which tends to magnify volatility. When these offsetting effects are taken into account, the finding that increased trade liberalization tends to reduce volatility for large regions, but increase volatility in smaller manufacturing centres still holds.