The prototypical parkinsonian tremor is a rest tremor of 4–6 Hz, appearing when the limb is fully supported and relaxed, characteristically asymmetric at onset and often remaining so, with the classic supination-pronation ("pill-rolling") quality in the hand (Bhatia et al., Mov Disord, 2018). It attenuates with voluntary movement and is exacerbated by cognitive load, contralateral movement, and emotional arousal — features that are diagnostically useful and phenomenologically distinctive. Beyond the hand, rest tremor may involve the leg, jaw, chin, and lips, while sparing the head and voice in a way that helps separate it from essential tremor.
Two further phenomenological points are worth emphasizing. First, tremor is not a universal feature: a substantial minority of patients have little or no tremor and present predominantly with bradykinesia and rigidity, so the absence of tremor never excludes the diagnosis (Postuma et al., Mov Disord, 2015). Second, the tremor's state-dependence is itself diagnostic — the way it emerges at rest, damps during a targeted movement, and can be provoked by contralateral motor tasks or serial mental arithmetic gives the bedside examiner reliable maneuvers to characterize it. These activation dynamics are not incidental; they are downstream signatures of the circuit behavior discussed below, which is why phenomenology and pathophysiology are best read together rather than as separate chapters.
Rest tremor, however, rarely occurs in phenotypic isolation. Many patients exhibit a postural and kinetic tremor as well, and the interplay of these components is part of why parkinsonian tremor is, as Fishman framed it, the most complex of the cardinal signs to characterize — its relationship to the other motor features is variable, with implications for diagnosis, prognosis, and treatment (Fishman, Mov Disord, 2008). The MDS clinical diagnostic criteria treat rest tremor as one supportive criterion among several, precisely because tremor phenotype alone is neither necessary nor sufficient for the diagnosis (Postuma et al., Mov Disord, 2015).
One phenomenon deserves particular attention because it is both clinically confusing and mechanistically informative. Re-emergent tremor is a postural tremor that appears not immediately on assuming a posture but after a variable latency of seconds, and that oscillates at the same frequency as the patient's rest tremor. The prevailing interpretation is that it is a continuation of the rest tremor — the same oscillator re-establishing itself once the limb settles into a new steady state — rather than a separate action tremor (Hallett, Parkinsonism Relat Disord, 2012; Belvisi et al., Parkinsonism Relat Disord, 2017).
This distinction matters at the bedside because re-emergent tremor visually mimics the postural action tremor of essential tremor, and the latency is often the discriminating feature. It matters mechanistically because re-emergent tremor is frequently therapy-refractory and has been linked to pathologic cortical oscillatory activity: recent work using simultaneous EEG-EMG has demonstrated distinct oscillatory signatures — including pathologic beta-band changes — that differentiate re-emergent from rest tremor and implicate the motor cortex in its maintenance (Dirkx & Bologna, Mov Disord, 2020; Mov Disord, 2024).
The central pathophysiological puzzle is that parkinsonian tremor behaves as though it is only loosely tethered to the disease's defining lesion. Whereas bradykinesia and rigidity scale with nigrostriatal dopaminergic denervation, tremor severity correlates poorly with dopaminergic imaging measures, and tremor does not respond as readily or predictably to levodopa as the other cardinal signs do (Hallett, Parkinsonism Relat Disord, 2012). Fishman catalogued these "paradoxical aspects" explicitly: patients with a pure rest tremor show dopaminergic deficits on radioligand imaging, yet tremor severity tracks those deficits weakly, and the pharmacologic response is complex rather than dose-linear (Fishman, Mov Disord, 2008).
A compelling resolution is that the dopaminergic response is itself heterogeneous and reflects individual differences in tremor circuitry. Comparing dopamine-responsive and dopamine-resistant tremor phenotypes, Dirkx and colleagues showed that dopaminergic medication reduces tremor by acting on pallidal and thalamic activity, and that interindividual variation in this pathway underlies the clinical heterogeneity — casting doubt on any simple "tremor is dopaminergic" account (Dirkx et al., Brain, 2019). In other words, the paradox is not that tremor is dopamine-independent, but that its dopaminergic control is partial, indirect, and patient-specific.
The anatomy that best accounts for these observations is a distributed one. Two partially separable cortico-subcortical circuits are implicated in Parkinson's motor signs: the striato-thalamo-cortical (STC) circuit, classically tied to bradykinesia and rigidity, and the cerebello-thalamo-cortical (CTC) circuit. Functional imaging during motor tasks shows differential engagement of these pathways between tremor-dominant and akinetic-rigid patients, with the CTC circuit implicated in the resting tremor phenotype (Prodoehl et al., Neuroscience, 2011). Consistent with this, single-unit recordings find that activity in the VIM nucleus of the thalamus — a cerebellar relay — is more coherent with tremor than basal-ganglia activity is, and that different body parts appear to have separate, non-phase-locked tremor generators (Hallett, Parkinsonism Relat Disord, 2012).
These strands converge on the influential "dimmer-switch" model: the basal ganglia act as a switch that triggers tremor episodes, while the cerebello-thalamo-cortical circuit sets tremor amplitude — the dimmer (Helmich, Curr Opin Neurol, 2020). The model elegantly reconciles the dopamine paradox (a dopaminergic basal-ganglia trigger, plus a largely non-dopaminergic cerebellar amplifier) with the efficacy of VIM-targeted therapies, and with the observation that the tremor's amplitude and its dopaminergic responsiveness are partly dissociable. Network-level electrophysiology reinforces the specificity of these loops: cerebello-cortical "fingerprints" differ measurably between essential tremor, Parkinson's tremor, and mimicked tremor, suggesting the circuits are not merely overlapping but distinguishable (Muthuraman et al., Brain, 2018). That distinguishability is a reminder that "tremor" is not one physiological entity but a family of oscillations produced by partly shared, partly distinct networks — a point with direct consequences for how any individual patient's tremor should be interpreted and treated.
Phenomenology and pathophysiology meet in the concept of motor subtypes. At one pole sits the tremor-dominant phenotype and its most striking variant, benign tremulous parkinsonism — prominent tremor with disproportionately mild bradykinesia and rigidity, slow progression, and often a limited dopaminergic response. Josephs and colleagues characterized this entity in a Mayo Clinic series of patients followed for at least eight years, establishing it as a recognizable, slowly evolving phenotype rather than an artifact of early disease (Josephs et al., Arch Neurol, 2006). At the other pole sits the postural-instability/gait-difficulty phenotype, with faster progression and greater non-tremor burden.
The subtype distinction is more than descriptive. It aligns with the circuit dissociation above — a tremor-dominant course implies a phenotype in which the cerebellar amplifier dominates and the dopaminergic-deficit-driven features lag — and it carries prognostic weight, since tremor-predominant presentations are broadly associated with slower progression. That association is one reason a prominent presenting tremor, though distressing, is not a marker of aggressive disease.
The distributed-network account is not merely of academic interest; it explains several otherwise puzzling clinical facts and shapes how tremor is treated. First, it accounts for why tremor can be the most therapy-refractory motor feature even when bradykinesia and rigidity respond well to dopaminergic therapy: a largely non-dopaminergic cerebellar amplifier is doing much of the work, so optimizing the dopaminergic regimen addresses only part of the circuit (Hallett, Parkinsonism Relat Disord, 2012). Second, it explains why interventions targeted at the cerebello-thalamo-cortical node — specifically the VIM thalamus — are so effective for tremor across tremor etiologies: thalamotomy, focused ultrasound, and VIM deep brain stimulation act on the shared final common pathway where tremor amplitude is generated, which is also why they help parkinsonian and essential tremor alike despite different upstream triggers (Helmich, Curr Opin Neurol, 2020).
Third, the model reframes a common counseling point: tremor and the bradykinetic-rigid syndrome can, to a degree, be treated as partly separable problems, which is why a patient may achieve good control of slowness and stiffness yet still request additional, tremor-specific treatment. Framing tremor as a distinct circuit-level phenomenon, rather than a symptom that should simply track the dopaminergic response, sets more accurate expectations and a more rational escalation pathway.
The phenomenological framework above is also the diagnostic one. The MDS tremor classification organizes tremor along two axes — clinical features (including activation state: rest versus action) and etiology — and this axis-based approach is what separates parkinsonian tremor from its principal mimics (Bhatia et al., Mov Disord, 2018). The key discriminations in practice: essential tremor is an action tremor without the rest component, other parkinsonian signs, or the asymmetry typical of Parkinson's; dystonic tremor is irregular, position-sensitive, and accompanied by dystonic posturing; and scans-without-evidence-of-dopaminergic-deficit (SWEDD) cases, historically confused with tremulous Parkinson's, show a normal dopaminergic scan and a different natural history.
Several questions remain genuinely open. The precise trigger by which basal-ganglia pathology recruits the cerebello-thalamo-cortical circuit is not established, and the dimmer-switch model, while influential, is a framework rather than a fully specified mechanism (Helmich, Curr Opin Neurol, 2020). Why some patients' tremor is dopamine-responsive and others' is not is only partly explained by the circuit differences described so far (Dirkx et al., Brain, 2019). The contribution of non-dopaminergic neuromodulatory systems — including serotonergic and noradrenergic pathways implicated in tremor by molecular imaging — is increasingly recognized but incompletely mapped, and may be part of why a purely dopaminergic account of tremor has always fit imperfectly; whether these systems represent viable, tremor-specific therapeutic targets remains an open and clinically consequential question. And the relationship between rest and re-emergent tremor, though best modeled as a continuation, still lacks a complete account of the cortical dynamics that gate the latency and the therapy-refractoriness (Mov Disord, 2024). The defensible summary is that parkinsonian tremor is a distributed-network phenomenon — triggered in the basal ganglia, amplified in cerebellar circuitry, only partially dopaminergic, and phenotypically heterogeneous — and that this framework, rather than a single lesion, is what a clinician should carry into the interpretation of any individual patient's tremor.