Continuous Glucose Monitoring (CGM): The Complete Guide

A continuous glucose monitor turns glucose from a once-in-a-while number into a continuous story. This guide explains how CGM works, how accurate it is, the metrics that matter, how to read your own data, and how the technology is used in and beyond diabetes care. It's educational, not medical advice.

What continuous glucose monitoring is

A CGM is a small wearable sensor, usually worn on the arm or abdomen, that measures glucose in the interstitial fluid just beneath the skin every few minutes, day and night. Instead of the isolated dots you get from a finger-stick meter, it produces a continuous line — a trace that shows not just how high or low glucose is, but when and how fast it changes. That shape is where a lot of the useful information lives, and it is invisible to a single blood test. For the deeper contrast, see HbA1c vs CGM.

How a CGM works

A thin filament under the skin senses glucose, a transmitter sends readings to a phone or reader, and software turns the stream into charts and alerts. Because it measures interstitial fluid rather than blood, there is a short lag between a CGM value and a finger-stick, especially when glucose is changing quickly. The lag between an interstitial CGM value and a finger-stick matters most when glucose is moving fast. Read more in How a CGM works.

How accurate is a CGM?

Modern CGMs are accurate enough for many day-to-day decisions, and their accuracy is often summarized with a metric called MARD. But accuracy varies by device, by glucose level, and in the first day of wear, and CGMs are not a perfect substitute for a lab test. See How accurate is a CGM? and CGM vs finger-stick.

The metrics that matter

A CGM produces a standard set of metrics. The most useful for most people are Time in Range (how much of the day glucose stays within target), the Glucose Management Indicator (GMI) (an estimated A1c from CGM data), and glucose variability (how much glucose swings). Each is worth understanding on its own:

You can convert between these numbers with our HbA1c & GMI calculator and check your day against the consensus targets with the Time-in-Range calculator.

Reading your CGM data

The single most useful CGM view is the Ambulatory Glucose Profile (AGP), a standardized report that overlays several days into one picture so patterns jump out. Learning to read it — the median line, the spread, the overnight window, the post-meal rises — turns raw data into action. See How to read your CGM data.

Choosing and wearing a CGM

Sensors differ in wear time, whether they need calibration, alerting, and cost. If you're deciding, start with How to choose a CGM and CGM sensors compared.

CGM in and beyond diabetes

CGM is well established in type 1 and increasingly in type 2 diabetes, including in primary care — see CGM for type 2 in primary care and CGM for prediabetes. Interest in using CGM for general metabolic insight is growing too; see CGM for people without diabetes. Whatever the setting, a CGM is a window, not a diagnosis — the readings are best interpreted with a clinician, and increasingly with software like Endobits that reads the shape of the curve as clinical decision support under physician oversight.

The full CGM library

Sources

American Diabetes Association — CGM & Time in Range.
Battelino T, et al. Clinical targets for CGM data interpretation (international consensus on Time in Range), Diabetes Care, 2019.
NIH / NIDDK — Continuous Glucose Monitoring.

Related hubs: Glossary · All resources