Retinoic acid has been previously proposed in the treatment of Alzheimer's disease (AD). Here, five transgenic mouse models expressing AD and frontotemporal dementia risk genes (i.e., PLB2 APP , PLB2 TAU , PLB1 Double , PLB1 Triple , and PLB4) were used to investigate if consistent alterations exist in multiple elements of the retinoic acid signaling pathway in these models. Many steps of the retinoic acid signaling pathway including binding proteins and metabolic enzymes decline, while the previously reported increase in RBP4 was only consistent at late (6 months) but not early (3 month) ages. The retinoic acid receptors were exceptional in their consistent decline in mRNA and protein with transcript decline of retinoic acid receptors  and ␥ by 3 months, before significant pathology, suggesting involvement in early stages of disease. Decline in RBP1 transcript may also be an early but not late marker of disease. The decline in the retinoic acid signaling system may therefore be a therapeutic target for AD and frontotemporal dementia. Thus, novel stable retinoic acid receptor modulators (RAR-Ms) activating multiple genomic and non-genomic pathways were probed for therapeutic control of gene expression in rat primary hippocampal and cortical cultures. RAR-Ms promoted the non-amyloidogenic pathway, repressed lipopolysaccharide induced inflammatory genes and induced genes with neurotrophic action. RAR-Ms had diverse effects on gene expression allowing particular RAR-Ms to be selected for maximal therapeutic effect. Overall the results demonstrated the early decline of retinoic acid signaling in AD and frontotemporal dementia models and the activity of stable and potent alternatives to retinoic acid as potential therapeutics.