In partnership with

For the business side of healthcare

Before the first meeting. Before the inbox chaos. Before a deal reshapes the room.

Healthcare Brew’s free newsletter delivers the business news shaping healthcare—covering hospital operations, pharma pricing, and everything in between—so you have the context behind the headlines when it counts.

Built for healthcare professionals whose work touches budgets, deals, and strategy.

Join 135K+ reading for free

Introduction

Dopamine has long been defined as the brain’s reward signal, the chemical that reinforces behaviors after success. That model has shaped decades of neuroscience, from learning theory to addiction research. But it leaves a critical gap: how does the brain adjust behavior while it’s happening, before the outcome is known?

A recent study published in Nature addresses that gap by examining dopamine activity during real-time, goal-directed movement. Using large-scale recordings across the striatum in mice navigating toward rewards, researchers were able to track how dopamine responds not just to expected rewards, but to the accuracy of ongoing behavior itself. The result is a shift in understanding: dopamine appears to function as a continuous feedback signal that evaluates whether actions are moving toward or away from a goal.

What the Study Found

The study revealed that dopamine activity tracks what the researchers term “trajectory error”, a continuous measure of how well an animal’s movement aligns with its goal. When movement direction and speed were correct, dopamine signals increased. When movement deviated from the optimal path, dopamine decreased, and the magnitude of this signal scaled with how incorrect the movement was.

Importantly, this signal was not isolated or rare. It appeared across the majority of recorded sites in the striatum, indicating that this is not a niche function but a widespread feature of dopamine signaling. The signal also persisted across different task conditions, whether driven by visual input or locomotion alone, suggesting it reflects a general computational principle rather than a task-specific artifact.

At the same time, dopamine continued to encode traditional reward-related information, expected value, reward history, and cue associations. However, these signals were separable from trajectory error, both in timing and anatomical distribution. This means the brain is not choosing between reward and error signals, it is running both simultaneously.

Mechanisms & Neuroscience

Dopamine as a Reinforcement Learning Signal

Dopamine has traditionally been modeled as a reward prediction error signal, the difference between expected and actual outcomes. This framework explains how the brain learns from experience by strengthening behaviors that lead to rewards.

This study expands that model by showing dopamine can also encode movement-dependent errors before outcomes occur. Using reinforcement learning simulations, researchers demonstrated that these trajectory error signals can emerge from the same computational framework, but only when the brain integrates multiple inputs, such as position, velocity, and sensory cues, into a unified state representation.

Trajectory Error: A New Type of Dopamine Signal

Trajectory error is fundamentally different from traditional reward signals. It is not binary, and it is not delayed. Instead, it is continuous, updating moment-to-moment as behavior unfolds.

This means the brain is not simply evaluating whether an action succeeded, it is constantly measuring how close the current action is to success. The signal integrates both direction and speed, effectively computing whether behavior is converging toward or diverging from a goal in real time.

This type of signal is far more useful for dynamic environments, where waiting for an outcome would be too slow to guide behavior effectively.

Neural Circuitry: How the Brain Separates Value and Error

One of the most important findings is that reward and error signals are not only functionally distinct, they are physically and temporally separated in the brain.

Trajectory error signals were strongest in anterior regions of the striatum, while reward-related signals were more prominent in medial regions. Additionally, reward signals appeared earlier in time, with error signals following shortly after (~0.3 seconds later).

This separation allows the brain to process both motivation and correction simultaneously without interference. Rather than blending into a single signal, dopamine is multiplexed, different components are layered together but remain distinguishable.

Sensory + Motor Integration: How the Brain Computes Error

Trajectory error is not derived from a single source of information. The study showed that dopamine can compute this signal using either visual input (external cues) or locomotion (internal movement signals), and often both.

This indicates that the brain integrates sensory and motor information into a unified estimate of performance. Even when one input is removed, the system can still function, suggesting redundancy and robustness in how error is computed.

This type of integration is essential for real-world behavior, where the brain must constantly reconcile internal intentions with external feedback.

Practical Applications for Brain Health

This discovery has direct implications for how we understand cognitive performance and behavior.

First, it suggests that the brain continuously monitors behavioral accuracy, not just outcomes. This could underlie how we detect subtle mistakes in real time, whether in movement, decision-making, or attention.

Second, it provides a framework for understanding behavioral flexibility. Efficient correction requires detecting deviations early, and this dopamine signal may be a core mechanism enabling rapid adjustment.

Third, disruptions in dopamine systems may impair not just motivation, but the ability to detect and correct errors. This has implications for conditions such as ADHD, addiction, and Parkinson’s disease, where behavioral control and adaptation are often compromised.

Finally, it reframes dopamine as part of a control system rather than a reward system, one that continuously aligns behavior with goals through ongoing feedback.

The Bottom Line

Dopamine is not just reinforcing what you did, it is evaluating what you are doing, in real time.

This suggests the brain is constantly measuring how closely your actions align with your goals, using dopamine not just to reward behavior, but to correct it as it unfolds.

Reference

Striatum-wide dopamine encodes trajectory errors separated from value
Nature
DOI: 10.1038/s41586-025-10083-1