In river restoration a key goal is to predict the nature of the communities that will develop following resto ration efforts and to forecast recovery rates and trajectories. Thus restoration ecology must be underpinned by know ledge of the pattern and process of ecological colonisation and succession. Here we investigate patterns of microinvertebrate (organisms passing through a 1 mm sieve but retained on a 63 μm one) colonisation and succession in a newly constructed urban river, the Jubilee River in Southern England, over 8 years. We also examine how the microinvertebrate assemblage in the Jubilee River differs from that in the adjacent River Thames and draw some lessons for river restoration. Total microinvertebrate abundance and richness in the River Jubilee differed significantly across years and community similarity (Jaccard's) was typically above 0.6 and increased towards the end of the study. Of the 17 microinvertebrate taxa collected in 2001, 14 were still found in 2007, suggesting that the tolerance model may be a major mechanism in the succession of microinvertebrate communities of new streams. Microinvertebrate assemblage composition was significantly different across years (ANOSIM) in the Jubilee River. NMDS showed that there were strong shifts between some years, particularly following high flows. Overall our results suggested that the colonisation and successional trajectory of the microinvertebrate assemblage was rapid, perhaps because of the close proximity of potential colonisers. Microinvertebrate abundance and mean taxon richness was significantly higher in the River Jubilee (4862 ± 326 m–2; 9.7 ± 0.2) than in the River Thames (1359 ± 120 m–2; 7.9 ± 0.4). The microinvertebrate assemblage that developed in the Jubilee was significantly different (ANOSIM) from that in the Thames and became more distinct over time suggesting that the local habitat template is a crucial determinant of the composition of the microinvertebrate assemblage. In this highly connected urban river we show that dispersal is unlikely to be a barrier to microinvertebrate colonisation and that the successional trajectory is rapid suggesting that recovery of microinvertebrate assemblages following effective river restoration should be quick. Longer term (> 3 years) data sets charting the development of ecosystems in new streams are rare, in both pristine and urban environments, and more are needed to verify our conclusions and further elucidate the relative importance of controls on successional trajectories operating at different scales.