ABSTRACTThe greatest unmet therapeutic need in Parkinson's disease is the development of treatment that slows the relentless progression of the neurodegenerative process. The concept of “disease modification” encompasses intervention types ranging from those designed to slow the underlying degeneration to treatments directed at regenerating or replacing lost neurons. To date all attempts to develop effective disease‐modifying therapy have failed. Many reasons have been proposed for these failures including our rudimentary understanding of disease pathogenesis and the assumption that each targeted mechanisms of disease apply to most patients with the same clinical diagnosis. Here we review all aspects of this broad field including general concepts and past challenges followed by a discussion of treatment approaches under the following 4 categories: (1) α‐synuclein, (2) pathogenic mechanisms distinct from α‐synuclein (most also potentially triggered by α‐synuclein toxicity), (3) non‐SNCA genetic subtypes of “PD,” and (4) possible disease‐modifying interventions not directly influencing the underlying PD pathobiology. We emphasize treatments that are currently under active clinical development and highlight a wide range of important outstanding questions and concerns that will need to be considered to advance the field of disease modification in PD. Critically, it is unknown whether the dysfunctional molecular pathways/organelles amenable to modification occur in a sequential fashion across most clinically affected individuals or manifest differentially in independent molecular subtypes of PD. It is possible that there is no “order of disruption” applicable to most patients but, rather, “type of disruption” applicable to subtypes dependent on unknown factors, including genetic variability and other causes for heterogeneity in PD. Knowing when (early vs late), which (eg, synaptic transmission, endosomal sorting and maturation, lysosomal degradation, mitochondrial biogenesis), and in whom (PD subtype) specific disrupted cell pathways are truly pathogenic versus compensatory or even protective, will be important in considering the use of single or combined (“cocktails”) putative disease‐modifying therapies to selectively target these processes. Beyond the current phase 2 or 3 studies underway evaluating treatments directed at oxidative stress (inosine), cytosolic Ca2+ (isradipine), iron (deferiprone), and extracellular α‐synuclein (passive immunization), and upcoming trials of interventions affecting c‐Abl, glucagon‐like peptide‐1, and glucocerebrosidase, it might be argued that further trials in populations not enriched for the targeted pathogenic process are doomed to repeat the failures of the past. © 2018 International Parkinson and Movement Disorder Society