Abstract : The over-arching objective of this research was to improve the quantitative basis for prediction of the occurrence of Coronal Mass Ejections (CMEs) and their impact on space weather. The most directly useful discovery, from the point of view of Space Weather forecasting, is that a CME associated with an erupting X-ray sigmoid is more likely to have flux rope ("magnetic cloud") structure at 1 AU than an arbitrarily-selected CME. On impacting the magnetosphere, the sigmoid-associated flux ropes in our database all produced at least "moderate" (Dst> 50 nT) geomagnetic storms, and several produced major storms. Hence, not only are sigmoidal active regions more likely to erupt, as we showed in earlier work (Canfield Hudson & McKenzie 1999), they are more likely to cause significant storms when they do. Also discovered was a signature of impending eruption in H-alpha imaging spectra of active regions, which could be used to predict the occurrence of eruptions a few hours in advance, whereas the sigmoid signature is more useful on time scales of days. These discoveries have been documented in 11 publications and 26 presentations.