Huntington's disease (HD), an adult-onset, dominantly-inherited neurodegenerative disease primarily affecting the striatum and deep cortical regions of the brain, is caused by a polyglutamine expansion in the N-terminal region of a 350 kD protein called huntingtin (htt). Huntingtin normally contains an N-terminal polyglutamine stretch of less than 36 amino acids, but if and when the stretch exceeds this number, the protein misfolds and aggregates into protein plaques called inclusion bodies. Concurrently, the protein becomes toxic. As polyglutamine repeat number increases, disease emerges at an earlier age while the protein tends to aggregate faster in vitro. To understand pathogenic mechanisms in HD at the cellular level, various molecular risk factors need to be linked to the fate of individual neurons, much like individual health characteristics in human patients can be linked to outcomes like recovery, illness, or death. In this dissertation, we employ automated microscopy technology to track thousands of individual neurons and precisely relate a range of molecular risk factors we can measure to outcomes we are interested in, thereby suggesting pathogenic mechanisms and pathways in which therapeutic intervention may prove effective. We establish critical limits on how the basic molecular characteristics of the disease, including htt levels, polyglutamine expansion length, and inclusion body formation, relate to neuronal death. We then evaluate the prognostic significance of various misfolded conformations of htt and discover that a conformation with a two-stranded, compact, hairpin configuration of the polyglutamine region is strongly predictive of death. This conformation is recognized by monoclonal antibody 3B5H10, which could serve in the future as an intermediate marker for neurotoxicity in high-throughput small molecule therapeutic screens. We then employ our newly acquired knowledge of HD cellular pathogenesis to therapeutically evaluate the effect of a series of small molecules and proteins impinging on several different putative pathogenic or therapeutic pathways. Several therapies we test show initial potential as therapeutics and should be investigated further.