There is currently no well-established biomarker for Parkinson's disease. The need to better diagnose the condition, define the subtypes of disease, and follow its course independent of any symptomatic drug effects is well-established. In this review, we will begin by reviewing the evidence for biological fluid biomarkers in Parkinson's disease. We will then touch upon the role of brain imaging in diagnosis and defining prognosis, as well as the value of studying motor phenotype and its potential applications for characterising Parkinson's disease subtypes with differing natural histories.