We studied the population dynamics of feral oilseed rape (Brassica napus) for 10 years (1993)(1994)(1995)(1996)(1997)(1998)(1999)(2000)(2001)(2002) in 3658 adjacent permanent 100 m quadrats in the verges of the M25 motorway around London, UK. The aim was to determine the relative importance of different factors affecting the observed temporal patterns of population dynamics and their spatial correlations. A wide range of population dynamics was observed (downward or upward trends, cycles, local extinctions and recolonizations), but overall the populations were not self-replacing (k < 1). Many quadrats remained unoccupied throughout the study period, but a few were occupied at high densities for all 10 years. Most quadrats showed transient oilseed rape populations, lasting 1-4 years.There were strong spatial patterns in mean population density, associated with soil conditions and the successional age of the plant community dominating the verge, and these large-scale spatial patterns were highly consistent from year to year. The importance of seed spilled from trucks in transit to the processing plant at Erith in Kent was confirmed: rape populations were significantly higher on the 'to Erith' verge than the 'from Erith' verge (overall mean 2.83-fold greater stem density). Quadrats in which k > 1 were much more frequent in the 'to Erith' verge, indicating that seed immigration can give the spurious impression of self-replacing population dynamics in time-series analysis.There was little evidence of a pervasive Moran effect, and climatic forcing did not produce widespread large-scale synchrony in population dynamics for the motorway as a whole; just 23% of quadrats had significant rank correlations with the mean time-series. There was, however, significant local spatial synchrony of population dynamics, apparently associated with soil disturbance and seed input. This study draws attention to the possibility that different processes may impose population synchrony at different scales. We hypothesize that synchrony in this system is driven by at least three processes: small-scale, local forcing caused by soil disturbance, intermediate-scale forcing as a result of seed input, and large-scale climatic forcing (e.g. winter rainfall) that affects the motorway as a whole.