Many pathways have been proposed as contributing to Huntington's disease (HD) pathogenesis, but generally the in vivo effects of their perturbation have not been compared with reference data from human patients. Here we examine how accurately mechanistically motivated and genetic HD models recapitulate the striatal gene expression phenotype of human HD. The representative genetic model was the R6/2 transgenic mouse, which expresses a fragment of the huntingtin protein containing a long CAG repeat. Pathogenic mechanisms examined include mitochondrial dysfunction; profiled in 3-nitropropionic acid-treated rats, 1-methyl-4-phenyl-1,2,3,6-tetrahydropyridine-treated mice, and PGC-1alpha knock-out mice; and depletion of brain-derived neurotrophic factor (BDNF) using heterozygous and forebrain-specific BDNF-knock-out mice (BDNF(HET), Emx-BDNF(KO)). Based on striatal gene expression, we find the BDNF models, both heterozygous and homozygous knock-outs, to be more like human HD than the other HD models. This implicates reduced trophic support as a major pathway contributing to striatal degeneration in HD. Because the majority of striatal BDNF is synthesized by cortical neurons, the data also imply that cortical dysfunction contributes to HD's hallmark effects on the basal ganglia. Finally, the results suggest that striatal lesions caused by mitochondrial toxins may arise via pathways different from those that drive neurodegeneration in HD. Based on these findings, we present a testable model of HD pathogenesis that, unlike most models, begins to account for regional specificity in human HD and the absence of such specificity in genetic mouse models of HD.