A 4-year study was initiated in 1997 to provide canola producers with information on the consequences of various rotational intervals with the use of new disease and herbicide-resistant canola varieties. The study was conducted at three locations in Alberta, Canada (Ellerslie, Strathmore, and Warburg). At each location, four canola rotations were established: continuous canola, and canola seeded in 1 of 2, 3, or 4 years. Canola varieties included the conventional varieties ‘AC Excel’ and ‘Quantum’, the glyphosate-resistant variety ‘Quest’, imidazolinone-resistant ‘45A71’, and a glufosinate-resistant hybrid, ‘Invigor 2153’. In the fourth year of the study, when canola was grown in all treatments, weed densities, weed species diversity, and evenness were determined preseeding and before and after in-crop herbicide application. Canola yield was greatest in the northern ecoregions of the Boreal Transition (Warburg) and Aspen Parkland (Ellerslie), and lowest in the Moist Mixed Grassland ecoregion (Strathmore). Weed populations increased and population diversity decreased and became less even where rotations were less diverse; in continuous canola and in the 1-in-4–year rotation. As expected, weed densities increased in poorly competitive, conventional canola varieties compared to herbicide-resistant varieties sprayed with their broader-spectrum herbicides. Where weed densities were high, variety/herbicide system became a critical factor affecting crop yield. Under these conditions, the herbicide-resistant varieties ‘Quest’ and ‘Invigor 2153’ typically outyielded the conventional varieties of ‘AC Excel’ and ‘Quantum’. Canola yield was highest when grown in a 1-in-3– or a 1-in-4–year rotation, although the 1-in-3–year rotation generally had lower weed densities, and allowed high-value canola to be grown more frequently in rotation.