Metabolic Timeline
How Long Does It Take to Get Into Ketosis?
Forty-eight to ninety-six hours for most people — and the reason that range is so wide has almost nothing to do with your metabolism and almost everything to do with how much glycogen you were carrying when you started.
The short answer
Hold carbohydrate under 30 g/day and blood ketones usually cross 0.5 mmol/L in 48–96 hours. Training, and a lower carbohydrate ceiling in week one only, shorten it. The weight you lose in those first days is water, not fat — and that distinction is the single biggest source of week-three disappointment in this diet.
What actually has to happen
Ketosis is not a switch that carbohydrate restriction flips. It is what happens once a specific fuel store runs out.
Your liver holds roughly 80–120 g of glycogen, and that store is what maintains blood glucose between meals. While it lasts, there is no metabolic reason for your liver to produce ketone bodies at any meaningful rate. Once it is substantially drained and not being refilled, fatty acid oxidation increases and the liver starts converting acetyl-CoA into beta-hydroxybutyrate — the ketone that matters and the one a blood meter measures.
So the question "how long until ketosis" is really the question "how long until my liver glycogen runs out", which is why the answer is so personal.
A realistic timeline
| Time | What is happening | Typical blood ketones |
|---|---|---|
| 0–12 h | Running on the last meal, then liver glycogen | 0.1–0.2 mmol/L |
| 12–24 h | Liver glycogen depleting; gluconeogenesis rising | 0.2–0.4 |
| 24–48 h | Fat oxidation up sharply; ketone production begins in earnest | 0.3–0.8 |
| 48–96 h | Nutritional ketosis established for most people | 0.5–1.5 |
| 1–3 weeks | Muscle and brain adapt to using ketones efficiently | 0.5–3.0 |
| 3–6 weeks | Keto-adaptation: performance recovers, ketones often fall as tissues use them better | 0.5–1.5 |
Note the last row, because it causes a great deal of confusion. Falling ketones after a month is usually a sign of adaptation, not failure. Your muscles have got better at consuming beta-hydroxybutyrate, so less of it accumulates in your blood for the meter to find. Being "in ketosis" and "having high blood ketones" stop being the same thing.
What makes your number faster or slower
Four things, in descending order of how much they matter:
- Starting glycogen. The dominant factor. Coming off a high-carbohydrate week with no training, you may have 500 g+ of total glycogen to work through. Coming off moderate eating and a hard session yesterday, far less.
- Training during the transition. The most effective lever you actually control. Glycogen is stored locally in the muscles that use it, so full-body work and longer moderate sessions deplete more than a heavy single-joint session does.
- Your carbohydrate ceiling. 20 g gets there faster than 50 g. Worth doing for the first week; not worth doing forever, because the stricter ceiling is also the one people abandon.
- Protein intake. Adequate protein is not a problem. Very high protein supplies gluconeogenic substrate and can blunt ketone production, so aim at your target rather than above it — see tracking keto macros.
Exogenous ketones are not a shortcut
Drinking ketone esters or salts raises the number on your meter within an hour. It does not accelerate glycogen depletion, and it does not make your liver start producing ketones — if anything, circulating exogenous ketones suppress endogenous production. You will read a higher number while being less far along. If your goal is the metabolic state rather than the reading, they are beside the point.
How to verify it, honestly
| Method | Measures | Verdict |
|---|---|---|
| Blood meter | Beta-hydroxybutyrate | The only one that measures what matters. >0.5 mmol/L = nutritional ketosis |
| Urine strips | Acetoacetate spillover | Useful for about two weeks, then misleading — you excrete less as you adapt |
| Breath meter | Acetone | Convenient, correlates loosely, no calibration standard |
| How you feel | — | Real signal, not a measurement |
The urine-strip row is responsible for an enormous amount of unnecessary anxiety. People watch the strips fade at week three and conclude the diet stopped working, when the fading is the adaptation. If you are going to measure, measure blood — and measure at the same time of day, because ketones follow a diurnal pattern and are typically lowest in the morning.
Keto flu is a sodium problem wearing a costume
Glycogen binds roughly three grams of water per gram. Depleting it flushes a large volume of fluid, and sodium leaves with the water.
The headache, fatigue, light-headedness, cramping and general flatness of days two to five track that far more closely than they track ketone levels. It is not a detox and it is not evidence of anything working.
Three to five grams of sodium a day — more than you think, and more than the diet will give you by accident — plus adequate potassium and magnesium resolves most of it within a day or two, and it does not slow the transition down. Details in keto electrolytes.
The five pounds in three days, and why it will hurt you later
Three grams of water per gram of glycogen means a full depletion moves several pounds off the scale in the first few days, irrespective of your calorie intake.
Two consequences, and the second is the one that ends diets:
- Week one tells you nothing about fat loss. Do not extrapolate from it.
- That water comes straight back the moment you eat carbohydrate. One high-carbohydrate meal can appear to undo three weeks of work overnight. It has not. But people believe the scale, feel they have failed, and quit.
Judge fat loss on a three-week trend of weekly averages, and discard week one from the calculation entirely.
Should you fast to speed it up?
It works — a 16 to 24 hour fast at the start depletes glycogen faster than low-carbohydrate eating with normal food volume, and usually pulls blood ketones above 0.5 mmol/L sooner. See keto and intermittent fasting.
Whether it is wise is a separate question. You would be stacking two unfamiliar adjustments in the same week, which is the combination people most often abandon. If your constraint is adherence rather than speed, change one variable. Two days' difference in your start date is worth very little; quitting in week one costs everything.
Where the transition usually goes wrong
Not the physiology — the accounting. The most common reason someone holds "under 30 g" for four days and sees no ketones is that they were not actually under 30 g.
Hidden carbohydrate concentrates in exactly the items nobody logs: two tablespoons of ketchup is about 9 g, barbecue sauce about 12 g, a cup of milk about 12 g, a handful of cashews about 9 g. Any one of those can consume a third to a half of the day's ceiling in a line item that never gets entered.
And since the largest error in carbohydrate tracking is portion estimation rather than the composition table, the accuracy of the tool matters more here than it does for general calorie counting — you cannot average a ketosis-ending meal away across the week. Worth knowing that only one app in this category, PlateLens, has a calorie-error figure that one independent laboratory measured and a second unrelated one then reproduced (±1.1% across two different weighed-meal sets); its carbohydrate figure of ±1.6% is the company's own claim rather than an independent measurement, and we distinguish the two. Comparison in best apps for keto.
Related
- Keto for beginners
- Keto electrolytes
- Keto macro calculator
- How to track keto macros
- Keto tracking mistakes
This article is general information, not medical advice. Ketogenic diets interact with insulin, sulfonylureas and SGLT2 inhibitors in ways that require dose adjustment and monitoring, and are not appropriate in pregnancy or with certain metabolic conditions. Talk to your clinician before starting one.