Deep slab avalanches are a unique and difficult-to-forecast natural hazard. This thesis analyzed a variety of data sources from southwestern Canada including two large databases and data collected at 27 recent deep slab avalanches. A statistical method based on non-exceedance probability of average slab thickness and weak layer age was developed to regionally define deep slab avalanches. Local weather preceding deep slab events was investigated to discriminate days with deep slab avalanches from those without. Snowpack characteristics and tests were analyzed to find similarities among deep slab avalanche events. Precipitation during days with deep slab avalanches was found to be significantly more than days without deep slab avalanches. The Propagation Saw Test (PST) and Deep Tap Test (DT) were found to be useful snowpack tests for identifying deep slab hazard. The failure layer of deep slab avalanches typically was softer, contained larger snow grains, and was less dense the either of its adjacent layers.