The initiation of movement, whether triggered by an internal (self-paced) or external stimulus (cued), relies on complex neural circuitry involving the pre-motor and motor cortex, basal ganglia, cerebellum, midbrain, and thalamus. The motor thalamus (mThal) integrates inputs from these structures to orchestrate the shift of cortical activity from a preparatory into an execution mode, resulting in action initiation. However, it is unclear whether the activity in the mThal that releases the motor program is different for externally cued and self-paced movements. Projections from the cerebellum and from the midbrain to the mThal have a critical role in cued movement initiation, while substantia nigra compacta (SNc) dopamine neurons, which send important projections to the basal ganglia, have been show to gate and invigorate self-paced actions. The loss of these neurons in Parkinson’s disease (PD) perturbs basal ganglia inputs to the mThal, leading to impairments in movement initiation. Interestingly, PD disrupts self-paced movements more than those triggered by external cues. We propose that self-paced movement initiation depends on the basal ganglia projections to the mThal, whereas external cues can engage the mThal through cerebellar and/or midbrain circuits to promote movement initiation. To test this hypothesis, we propose a novel behaviour task where mice initiate the same action in a cued or a self-paced manner, combined with recordings and optogenetic manipulation of input-defined mThal populations. We expand this approach to a mouse model of PD to further test if our hypothesis can explain why the use of external cues has a beneficial effect in PD. Our ground-breaking experiments will advance the fundamental understanding of the neural circuits that support self-paced and cued movement initiations, while establishing a framework to understand the positive effect of cues in PD.