# How many ATP are produced in glycolysis?

> Glycolysis makes 4 ATP and spends 2, a net gain of 2 ATP per glucose. See which steps spend and make ATP, what its 2 NADH are worth with and without oxygen, and why glycogen gives 3.

Glycolysis produces **4 ATP** per glucose molecule but spends 2 ATP to get started, so its net yield is **2 ATP per glucose**. The 4 is the gross, or total, figure, made in the second half of the pathway; the 2 is what the cell keeps. Glycolysis also turns one glucose into 2 pyruvate and 2 NADH, and neither counts toward its direct ATP total.

Both numbers appear in textbooks and exam keys, and both are right once the question says which one it means. Oxygen does not change them. What changes is the value of the products: with oxygen the 2 NADH are worth about 3 or 5 more ATP, and in fermentation they are worth none. The starting point matters too. Glucose taken from glycogen nets 3 ATP, and red blood cells often keep fewer than 2.

## Why does glycolysis make 4 ATP but net only 2?

Glycolysis is a sequence of ten reactions in the cytosol, and only four of them involve ATP. Two early reactions spend one ATP each, and two late reactions make one ATP each, but the late reactions run twice for every glucose. That asymmetry gives 4 ATP made, 2 spent and 2 gained.

The first half is the investment phase. Hexokinase spends one ATP to turn glucose into glucose 6-phosphate, which traps the sugar in the cell. Phosphofructokinase-1 (PFK-1) spends a second to make fructose 1,6-bisphosphate, the irreversible, rate-limiting step. Aldolase then splits this sugar in two, and triose phosphate isomerase converts one half into the other, so two molecules of glyceraldehyde 3-phosphate enter the second half.

That second half is the payoff phase, and each of its five reactions happens twice per glucose. Phosphoglycerate kinase transfers a phosphate from 1,3-bisphosphoglycerate to ADP, and pyruvate kinase does the same from phosphoenolpyruvate. Each makes one ATP per three-carbon molecule, or 2 per glucose, for a gross total of 4. This direct transfer of a phosphate from a substrate to ADP is called substrate-level phosphorylation.

![Hexokinase and PFK-1 each use one ATP per glucose, while phosphoglycerate kinase and pyruvate kinase each make two ATP per glucose](/images/charts/glycolysis-atp-accounting.webp '**Figure 1. The four ATP steps of glycolysis.** ATP spent or made per glucose, shown as a positive magnitude; the labels say which steps consume ATP. Source: [Chaudhry and Varacallo (2023)](https://www.ncbi.nlm.nih.gov/books/NBK482303/).')

Table 1 places those four steps in the whole pathway. Summed over all ten reactions, glycolysis converts glucose, 2 NAD⁺, 2 ADP and 2 inorganic phosphate into 2 pyruvate, 2 NADH, 2 H⁺, 2 ATP and 2 water, and the ATP in that equation is the net figure.

  <caption><strong>Table 1. ATP and NADH at each step of glycolysis.</strong> Counts per glucose; negative values are ATP spent. Steps 6 to 10 run twice per glucose because step 4 splits the sugar in two. Sources: Alberts et al. (2002), Chaudhry and Varacallo (2023).</caption>
  <thead>
    <tr>
      <th scope="col">Step</th>
      <th scope="col">Enzyme</th>
      <th scope="col">Reaction</th>
      <th scope="col">ATP</th>
      <th scope="col">NADH</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>1</td>
      <td>Hexokinase</td>
      <td>Glucose to glucose 6-phosphate</td>
      <td>−1</td>
      <td>0</td>
    </tr>
    <tr>
      <td>2</td>
      <td>Phosphoglucose isomerase</td>
      <td>Glucose 6-phosphate to fructose 6-phosphate</td>
      <td>0</td>
      <td>0</td>
    </tr>
    <tr>
      <td>3</td>
      <td>Phosphofructokinase-1</td>
      <td>Fructose 6-phosphate to fructose 1,6-bisphosphate</td>
      <td>−1</td>
      <td>0</td>
    </tr>
    <tr>
      <td>4</td>
      <td>Aldolase</td>
      <td>Fructose 1,6-bisphosphate to dihydroxyacetone phosphate and glyceraldehyde 3-phosphate</td>
      <td>0</td>
      <td>0</td>
    </tr>
    <tr>
      <td>5</td>
      <td>Triose phosphate isomerase</td>
      <td>Dihydroxyacetone phosphate to glyceraldehyde 3-phosphate</td>
      <td>0</td>
      <td>0</td>
    </tr>
    <tr>
      <td>6</td>
      <td>Glyceraldehyde 3-phosphate dehydrogenase</td>
      <td>Glyceraldehyde 3-phosphate to 1,3-bisphosphoglycerate</td>
      <td>0</td>
      <td>+2</td>
    </tr>
    <tr>
      <td>7</td>
      <td>Phosphoglycerate kinase</td>
      <td>1,3-Bisphosphoglycerate to 3-phosphoglycerate</td>
      <td>+2</td>
      <td>0</td>
    </tr>
    <tr>
      <td>8</td>
      <td>Phosphoglycerate mutase</td>
      <td>3-Phosphoglycerate to 2-phosphoglycerate</td>
      <td>0</td>
      <td>0</td>
    </tr>
    <tr>
      <td>9</td>
      <td>Enolase</td>
      <td>2-Phosphoglycerate to phosphoenolpyruvate</td>
      <td>0</td>
      <td>0</td>
    </tr>
    <tr>
      <td>10</td>
      <td>Pyruvate kinase</td>
      <td>Phosphoenolpyruvate to pyruvate</td>
      <td>+2</td>
      <td>0</td>
    </tr>
    <tr>
      <td>Total</td>
      <td></td>
      <td>4 ATP made, 2 spent</td>
      <td>+2 net</td>
      <td>+2</td>
    </tr>
  </tbody>

Step 6 is also where the [2 NADH of glycolysis](/guides/how-many-nadh-are-produced-in-glycolysis) form. Its oxidation of glyceraldehyde 3-phosphate is what creates the high-energy phosphate bond that step 7 cashes in, so steps 6 and 7 together account for the net yield of 2 ATP and 2 NADH.

## How much ATP is glycolysis worth with and without oxygen?

Glycolysis makes the same net 2 ATP whether or not oxygen is present. Oxygen decides what the 2 NADH are worth afterwards, which is why answers to "total ATP in glycolysis" range from 2 to about 7.

With oxygen, the electrons from glycolytic NADH pass into the [mitochondria](/guides/how-many-mitochondria-are-in-a-cell). The inner mitochondrial membrane does not let NADH through, so one of two shuttles carries the electrons instead. The malate-aspartate shuttle regenerates NADH inside the mitochondrion, where its electrons enter the respiratory chain at complex I. The glycerol 3-phosphate shuttle delivers them further along, bypassing complex I, so they pump fewer protons.

In Hinkle's review of measured P/O ratios, electrons from NADH-linked substrates support about 2.5 ATP per pair and electrons from succinate, which also bypass complex I, about 1.5. The 2 NADH are therefore worth about 5 ATP through the malate-aspartate shuttle and about 3 through the glycerol 3-phosphate shuttle, bringing glycolysis to about 7 or 5 ATP per glucose.

Older textbooks, which credit each NADH with 3 ATP, or 2 after the glycerol 3-phosphate shuttle, give 8 or 6. The pyruvate carries the rest of glucose's energy into the [Krebs cycle](/guides/how-many-atp-are-produced-in-the-krebs-cycle) and [oxidative phosphorylation](/guides/how-many-atp-are-produced-in-oxidative-phosphorylation), for a total of about 30 to 32 ATP from [cellular respiration](/guides/how-many-atp-are-produced-in-cellular-respiration).

Without oxygen, the NADH cannot unload its electrons in the mitochondrion. In [fermentation](/guides/how-many-atp-are-produced-in-fermentation), the cell uses it to reduce pyruvate to lactate, as in hard-working muscle, or to make ethanol, as in yeast. That regenerates the NAD⁺ step 6 needs, but the NADH is spent for no ATP, and the cell keeps only the 2 net ATP.

  <caption><strong>Table 2. ATP from glycolysis of one glucose, by the fate of its NADH.</strong> Modern accounting uses about 2.5 ATP per NADH whose electrons enter at complex I and about 1.5 per NADH whose electrons bypass it; older accounting uses 3 and 2. Pyruvate oxidation is not included. Sources: Alberts et al. (2002), McKenna et al. (2006), Hinkle (2005).</caption>
  <thead>
    <tr>
      <th scope="col">Fate of the 2 NADH</th>
      <th scope="col">Net ATP made directly</th>
      <th scope="col">ATP from 2 NADH</th>
      <th scope="col">Total</th>
    </tr>
  </thead>
  <tbody>
    <tr>
      <td>Malate-aspartate shuttle, with oxygen</td>
      <td>2</td>
      <td>About 5</td>
      <td>About 7</td>
    </tr>
    <tr>
      <td>Glycerol 3-phosphate shuttle, with oxygen</td>
      <td>2</td>
      <td>About 3</td>
      <td>About 5</td>
    </tr>
    <tr>
      <td>Fermentation to lactate or ethanol</td>
      <td>2</td>
      <td>0</td>
      <td>2</td>
    </tr>
    <tr>
      <td>Older accounting, with oxygen</td>
      <td>2</td>
      <td>4 or 6</td>
      <td>6 or 8</td>
    </tr>
  </tbody>

## When does glycolysis yield more or less than 2 ATP?

The net 2 ATP assumes that glycolysis starts from free glucose and runs every step as written. Two common cases break those assumptions in opposite directions.

When a cell breaks down glycogen for its own use, as skeletal muscle does, glycogen phosphorylase cuts off glucose units using inorganic phosphate rather than ATP, releasing glucose 1-phosphate. Phosphoglucomutase converts it to glucose 6-phosphate at no energy cost. The sugar enters glycolysis already past step 1, and the hexokinase ATP is never spent.

By the same step count, each glucose unit released this way nets 3 ATP: 4 made and only 1 spent, at PFK-1. Muscle keeps its glycogen for rapid ATP generation, so a sprinting muscle running on its own glycogen gets half as much again from each glucose unit as one importing glucose from the blood.

[Red blood cells](/guides/how-many-red-blood-cells-are-in-the-human-body) go the other way. They have no mitochondria, so glycolysis ending in lactate is their only source of ATP. They also divert part of their 1,3-bisphosphoglycerate through the Rapoport-Luebering shunt, which makes 2,3-bisphosphoglycerate (2,3-BPG), the molecule that binds [hemoglobin](/guides/how-many-hemoglobin-molecules-in-a-red-blood-cell) and helps it release oxygen to tissues.

The 2,3-BPG is later dephosphorylated to 3-phosphoglycerate and rejoins the pathway, but the detour skips phosphoglycerate kinase, which Cho and colleagues describe as an energetic cost. Each three-carbon molecule that takes it gives up one ATP, so a glucose whose two halves both take the shunt nets 0 ATP, and a red blood cell's average yield falls below 2 in proportion to how much of its glycolysis goes through 2,3-BPG.

## Gross, net and what the pathway is worth

The three common answers measure different things. Four is what glycolysis makes, 2 is what it keeps after paying for its first steps, and about 5 to 7 is its worth to a cell that can burn the NADH with oxygen. Glycogen raises the kept share to 3, and the red blood cell's detour for oxygen delivery lowers it. The ATP itself always comes from the same two enzymes, phosphoglycerate kinase and pyruvate kinase.