Objective To evaluate a modified extended Timed Up and Go (extended-TUG) assessment against a panel of validated clinical assessments, as an indicator of Parkinson's disease (PD) severity and cognitive impairment. Methods Eighty-seven participants with idiopathic PD were sequentially recruited from a Movement Disorders Clinic. An extended-TUG assessment was employed which required participants to stand from a seated position, walk in a straight line for 7 m, turn 180° and then return to the start, in a seated position. The extended-TUG assessment duration was correlated to a panel of clinical assessments, including the Unified Parkinson's Disease Rating Scale (MDS-UPDRS), Quality of Life (PDQ-39), Scales for Outcomes in Parkinson's Disease (SCOPA-Cog), revised Addenbrooke's Cognitive Index (ACE-R) and Barratt's Impulsivity Scale 11 (BIS-11). Results Extended-TUG time was significantly correlated to MDS-UPDRS III score and to SCOPA-Cog, ACE-R (p < 0.001) and PDQ-39 scores (p < 0.01). Generalized linear models determined the extended-TUG to be a sole variable in predicting ACE-R or SCOPA-Cog scores. Patients in the fastest extended-TUG tertile were predicted to perform 8.3 and 13.4 points better in the SCOPA-Cog and ACE-R assessments, respectively, than the slowest group. Patients who exceeded the dementia cut-off scores with these instruments exhibited significantly longer extended-TUG times. Conclusions Extended-TUG performance appears to be a useful indicator of cognition as well as motor function and quality of life in PD, and warrants further evaluation as a first line assessment tool to monitor disease severity and response to treatment. Poor extended-TUG performance may identify patients without overt cognitive impairment form whom cognitive assessment is needed.