A European haplotype of common reed is expanding in roadside and agricultural ditches in Eastern Canada, forming monospecific linear colonies that often alternate with cattails (Typha sp.). No successful seedling establishment having been reported for this region, this expansion has been entirely attributed to accidental transportation of plant fragments followed by rapid clonal growth. We used aerial photographs, field observations, and mesocosm experiments to assess expansion rates in ditches and to determine the effect of competitive interactions with cattails on reed dynamics. A reconstruction of reed invasion over the last 15 years from remote sensing shows the frequent establishment of new reed patches and generally high rates of population increase. Field observations and manipulations over 3 years show that reed is faster than cattail at colonizing unvegetated sites. When two colonies meet, common reed is always gaining ground, at the expense of cattail, but this progression varies annually, probably in response to climatic conditions. A controlled competition experiment at high plant density shows that both common reed and cattail density and biomass are negatively affected. The detrimental effect of competition, however, is stronger on cattail. All these observations demonstrate the competitive superiority of common reed in ditch conditions. Finally, we bring new evidence of successful common reed establishment through seed germination and seedling establishment in roadside ditches. While it is possible that the contribution of seedlings may have been overlooked in the past, we believe it is more likely a rather recent phenomenon resulting from recent climate warming.