Course contentsShow
← React
Lesson 1433 of 2,39636. Concurrent ReactFree lesson

Urgent vs Transition Updates Prioritization

How React prioritizes user input and urgent updates over transitions to keep the UI responsive.

Urgent vs Transition Updates Prioritization

What you'll learn: How React treats user input as urgent while allowing transitions to run in the background without blocking the interface.

The Two-Lane System

React's Concurrent rendering introduces a prioritization system with two lanes: urgent updates and transition updates. Think of it like a hospital emergency room—critical injuries get immediate attention (urgent), while routine check-ups wait their turn (transitions).

Urgent Updates

By default, all state updates are urgent. These include:

  • Text input typing
  • Button clicks
  • Form submissions
  • Any direct user interaction feedback

React processes these immediately to keep your app feeling responsive. If a user types in a search box, they expect to see their keystrokes appear instantly.

Transition Updates

Updates wrapped with startTransition are marked as non-urgent. These typically involve:

  • Rendering expensive filtered lists
  • Complex visual updates
  • Data transformations
  • Navigation to heavy components

React can interrupt transition updates if an urgent update arrives. When you type the next character while a filter is still rendering, React pauses the filter work, handles your keystroke, then resumes the filtering.

How Prioritization Works

When both types of updates compete:

  1. Urgent updates always win — React interrupts transition rendering
  2. Transitions yield the main thread — Your UI stays responsive
  3. Transitions resume when urgent work finishes — No updates are lost

This is why isPending exists—it tells you when a transition is happening so you can show appropriate feedback without the entire UI freezing.

Key Takeaway: React's priority system ensures urgent user interactions never wait for expensive transitions, creating interfaces that feel instant even during heavy computation.