A field experiment was conducted to investigate the causes of poor canola growth through surface-retained wheat stubble. The experiment was designed to separate the role of biochemical factors such as stubble phytotoxicity, nitrogen (N) immobilisation and disease incidence from the physical effects of the stubble including changes in the seedbed microclimate on crop growth and yield, as these had been previously identified as possible mechanisms for the poor growth. Treatments included: two stubble cultivars, three levels of decomposition, two levels of nitrogen and inert plastic mulch. A novel approach involving leachates collected from intact field cores with a rainfall simulator was also used to assess the phytotoxic impacts of the stubble. Surface-retained wheat stubble (5 t/ha) reduced the canola vegetative biomass by 46% and yield by 26%, consistent with previous field observations. There was no impact of stubble variety or level of stubble decomposition on canola growth and yield, and an inert plastic mulch designed to simulate the physical effects of the wheat stubble had a similar effect on growth and yield as wheat stubble. Leachates collected from intact field cores through surface stubble also had no impact on canola germination or radicle elongation in petri-dishes in the absence of soil, nor on emergence of canola in pots of soil. Together, these observations suggest phytotoxicity was not contributing to the poor growth. N-immobilisation was evident in stubble-retained treatments although tissue N concentrations were adequate in all treatments and the growth limitation could not be overcome with added N. The principal mechanism of growth reduction was associated with the physical impact of the stubble including a reduction in photosynthetically active radiation, the red:far red ratio of incident light under the stubble, and the temperature above the stubble layer. These effects led to elongated hypocotyls, reduced investment in early leaf aroot growth, delayed emergence and slower leaf area development as well as an increase in seedling disease complex, effects apparent in both wheat stubble and inert plastic mulch treatments. The results suggest that physical, rather than biochemical factors are the main cause of poor growth of canola in surface-retained wheat stubble, and that much of the impact could be avoided if the stubble were moved away from directly above the emerging seedlings.