Development of a winter-hardy spring wheat breeding platform could increase the gain in selection per year over traditional winter wheat breeding programs. To make use of spring wheat being able to produce three generations per year, an indoor cold tolerance screen using chlorophyll fluorescence (Fv/Fm) and visual assessment two weeks after freezing as evaluation parameters was developed. Evaluation of Ontario-adapted winter and spring wheat varieties demonstrated that the test was able to differentiate between winter and spring wheat. Specific varieties from this data set were used to develop an indoor freezing survival index (IFSI) to normalize data for effective ranking of germplasm in further experiments. Indepth analysis of a Froid (winter) x Siete Cerros (spring) wheat population using molecular markers indicated that a significant level of cold tolerance is preserved when the Vrn-B1 spring allele is used compared to the Vrn-A1 allele. Generation means analysis of the same cross indicated that the cold tolerance was due to additive genetic effects. Multiple populations with at least one spring parent were advanced to the F3:4 generation. IFSI analysis indicated that several lines from the populations had cold tolerance similar to Ontario-adapted winter wheats and better than several winter barley varieites. Further testing found a significant 5% improvement in cold tolerance was noted for spring wheat varieties treated with Cruiser Maxx seed treatment. Finally, a Norstar (winter) x Bergen (spring) doubled haploid wheat population was analysed and a significant correlation to LT50 data from an independent laboratory validated the methods used in these experiments. In a separate experiment, multiple indicies calculated from spectral reflectance measurements taken on the Ontario winter wheat performance trial at Elora and Harriston in 2008-09 were found to be significantly correlated to winter survival ratings. Fall reflectance measurements indicate non-random plant density or vigour effects in the trials. To adjust winter survival ratings accordingly, linear and non-linear approaches were used and found the non-linear model to be statistically superior. Large differences between locations illustrated that for complete modelling of winter survival, more data from locations of differing soil types, plant density and plant growth stage is required.