Total Blocking Time (TBT)

Total Blocking Time (TBT) estimates how much main-thread work could delay a user's input during a defined load interval. In common Lighthouse-style measurement, tasks longer than 50 milliseconds contribute only the amount beyond that threshold. For example, a 120-millisecond long task contributes 70 milliseconds of blocking time. The reported total reflects a particular lab run, test environment, and measurement window.

A local service website may show a large TBT when it loads heavy booking widgets, third-party advertising scripts, or complex animations before visitors can interact. A developer can inspect a performance trace, identify long tasks, reduce unnecessary JavaScript, and retest after changes. Simply compressing images will not necessarily resolve JavaScript execution blocking the main thread.

TBT is not a field-measured Core Web Vital, and it should not be confused with Interaction to Next Paint (INP), which reflects real interaction responsiveness. Google’s web.dev guidance treats TBT as a lab proxy that can help investigate interaction issues, not a direct guarantee of better real-user outcomes. Compare repeat tests under similar conditions and examine actual-user metrics where available. This term is about main-thread load responsiveness, not the full loading time or server response time of a website.