Parkinson's disease (PD) is an age-related, progressive, multisystem neurodegenerative disorder resulting in significant morbidity and mortality, as well as a growing social and financial burden in an aging population. The hallmark of PD is loss of dopaminergic neurons of the substantia nigra pars compacta, leading to bradykinesia, rigidity and tremor. Current pharmacological treatment is therefore centred upon dopamine replacement to alleviate symptoms. However, two major problems complicate this approach: (i) motor symptoms continue to progress, requiring increasing doses of medication, which result in both short-term adverse effects and intermediate- to long-term motor complications; (ii) dopamine replacement does little to treat non-dopaminergic motor and non-motor symptoms, which are an important source of morbidity, including dementia, sleep disturbances, depression, orthostatic hypotension, and postural instability leading to falls. It is critical, therefore, to develop a broader and more fundamental therapeutic approach to PD, and major research efforts have focused upon developing neuroprotective interventions. Despite many encouraging preclinical data suggesting the possibility of addressing the underlying pathophysiology by slowing cell loss, efforts to translate this into the clinical realm have largely proved disappointing in the past. Barriers to finding neuroprotective or disease-modifying drugs in PD include a lack of validated biomarkers of progression, which hampers clinical trial design and interpretation; difficulties separating symptomatic and neuroprotective effects of candidate neuroprotective therapies; and possibly fundamental flaws in some of the basic preclinical models and testing. However, three recent clinical trials have used a novel delayed-start design in an attempt to overcome some of these roadblocks. While not examining markers of cell loss and function, which would determine neuroprotective effects, this trial design pragmatically tests whether earlier versus later intervention is beneficial. If positive (i.e. if an earlier intervention proves more effective), this demonstrates disease modification, which could result from neuroprotection or from other mechanisms. This strategy therefore provides a first step towards supporting neuroprotection in PD. Of the three delayed-start design clinical trials, two have investigated early versus later start of rasagiline, a specific irreversible monoamine oxidase B inhibitor. Each trial has supported, although not proven, disease-modifying effects. A third delayed-start-design clinical trial examining potential disease-modifying effects of pramipexole has unfortunately reportedly been negative according to preliminary presentations. The suggestion that rasagiline is disease modifying is made all the more compelling by in vitro and PD animal-model studies in which rasagiline was shown to have neuroprotective effects. In this review, we examine efforts to demonstrate neuroprotection in PD to date, describe ongoing neuroprotection trials, and critically discuss the results of the most recent delayed-start clinical trials that test possible disease-modifying activities of rasagiline and pramipexole in PD.