Potatoes and Blood Sugar: White, Sweet, Baked, Fried, and Cooled With Diabetes

Potatoes are one of the first foods people worry about after a diagnosis. They are cheap, filling, and part of nearly every cuisine, and they also show up on lists of foods to avoid. The honest answer sits in the middle. Potatoes are a concentrated source of carbohydrate, and how much you eat, how you cook them, and what else is on the plate all change the picture. This guide walks through the numbers, the glycemic index research, the cooling trick, fries, sweet potatoes, and how to test potatoes on your own body.

Quick answer: Many people with diabetes can eat potatoes, but they are not a free vegetable. A medium baked russet potato (173 g) with skin has about 37.0 g of carbohydrate and 4.0 g of fiber, and a half-cup of boiled potato has about 15.7 g. Portion size does most of the work. Cooking and cooling methods may matter somewhat, and the research is smaller than social media suggests. Your own glucose readings are the final test. This is general education, not a personal meal plan.

How much carbohydrate is in a potato?

The figures below come from USDA FoodData Central, the government nutrition database, as reproduced by public mirrors of the SR Legacy dataset. Per-100 g values are what the database lists. The per-serving rows are my arithmetic, using the serving weights the database gives (for example 173 g for a medium baked potato), so treat them as estimates. Real potatoes vary a lot in size, and a restaurant “large” baked potato can weigh well over 300 g.

A few things stand out. A single medium baked potato is already in the high 30s of carbohydrate grams, which is about the same as a cup of cooked rice. Our guide to white rice and blood sugar lists a cup of cooked white rice at about 44.5 g, so a medium potato is in the same neighborhood, with more fiber. A cup of mashed potatoes made with whole milk has about 37 g of carbohydrate by the USDA label-style entry, and that is before gravy. Ten oven-heated fries land near 19 g, and most people eat more than ten.

The other thing to notice is how little the carbohydrate per 100 g moves between cooking methods. Baked, boiled, and mashed potatoes all sit between roughly 17.6 and 21.6 g per 100 g. Cooking does not turn a potato into a different food. What changes is the serving size people eat, the fat and salt that get added, and, to some degree, how fast the starch is digested. The next sections cover that last part.

Preparation methods compared

This table is a summary of what the sources below report, not a ranking. The studies used different potato varieties, portion sizes, and test volunteers, so no preparation method wins every time.

What does the glycemic index say about potatoes?

The glycemic index (GI) ranks foods by how quickly they raise blood glucose compared with a reference food in healthy volunteers. It is a rough guide and does not predict your personal response. Potatoes are a famous example of why: the same vegetable can score anywhere from the 50s to over 100 depending on the study.

The first five rows come from a Harvard Health Publications table that is based on the 2002 edition of the International Tables of Glycemic Index. I could not open the 2021 International Tables in full, so I am not quoting its potato values here. The newer edition lists over 4,000 foods and sorts values by data quality, which is a good sign that single numbers should be held loosely.

The last two rows come from a study at Lund University in Sweden. Fourteen healthy adults ate meals with 50 g of available carbohydrate from fries, boiled potatoes, or white bread, and the researchers measured capillary glucose for three hours. Fries produced a significantly lower response than boiled potatoes. A GI above 100 is possible when white bread, not pure glucose, is the reference, which is one reason these numbers do not line up neatly with the Harvard table.

What should you take from this? That boiled potatoes can come out low in one study and very high in another. The International Tables authors say outright that more testing is needed. The practical message is that a potato is a reliably carbohydrate-heavy food and an unreliably predictable one, so your own meter or CGM data are worth more than any table. Our guide to reading CGM data and post-meal glucose tracker can help you set that up.

Glycemic load, and what Harvard says

Glycemic load multiplies the glycemic index by the carbohydrate in a serving, so it accounts for how much you actually eat. In the Harvard Health table, a 150 g serving of baked russet potato has a glycemic load of 26, compared with 14 for boiled white potato and 17 for sweet potato. The Harvard T.H. Chan School of Public Health notes that potatoes are left out of its Healthy Eating Plate vegetable group because of their high glycemic load, and that large cohort studies have linked higher potato and fry intake to weight gain and, in women, to higher diabetes risk. In the 2011 Mozaffarian analysis of about 120,000 people followed for up to 20 years, higher fry intake was linked to about 3.4 extra pounds every four years, and higher baked or mashed potato intake to about 1.3 extra pounds.

These are observational findings. They show an association, not proof that potatoes caused the outcome, and they describe populations, not the person at your table. They are still a useful reason to treat potatoes as a starch you measure rather than a vegetable you eat without thinking, and Harvard suggests whole intact grains or beans as starchy side dishes instead.

Does cooling a potato lower its blood sugar effect?

When a cooked potato cools, some starch can turn into resistant starch, which is digested more slowly. That is the idea behind eating potato salad or reheated leftovers. The human evidence is real but small. One randomized crossover study with 8 participants compared a freshly cooked mashed potato meal (203 g boiled potato and 25 g butter) with the same meal chilled for 66 hours and microwave reheated. Post-meal glucose was not significantly different. Insulin was: the reheated meal had a 24% lower incremental area under the curve over the first two hours and a 21% lower insulin peak. The authors suggested better liver insulin sensitivity as a possible reason and described the findings as preliminary.

So the fair summary is this. Cooling may change how your body handles the starch, the glucose effect was not shown, and eight people is a very small group. It costs nothing to try, since leftovers already exist in most kitchens. If you test it, keep the portion the same and compare readings from a freshly cooked potato and a cooled-and-reheated one on separate days.

Skin, fiber, and potassium

Potato skin is where much of the fiber lives. USDA lists baked russet with skin at 2.3 g of fiber per 100 g, compared with 1.5 g for the flesh alone. A separate USDA entry for baked skin alone lists 7.9 g of fiber per 100 g, or about 4.6 g for one 58 g skin. That entry also shows 46.1 g of carbohydrate per 100 g, which is higher than the potato itself, probably because the skin loses water during baking. I would not read that as skin being a carbohydrate bomb. The practical point is that eating the skin adds fiber, but it does not remove the carbohydrate in the flesh.

Potatoes are also a good potassium source, with about 550 mg per 100 g for baked russet with skin. That is helpful for many people. If you have kidney disease or take medicines that raise potassium, ask your care team before leaning on potatoes.

What about french fries?

Oven-heated frozen fries have about 25.6 g of carbohydrate and 5.5 g of fat per 100 g in the USDA entry, so ten fries (76 g) come to about 19.4 g of carbohydrate and 4.2 g of fat. Fast-food and restaurant fries differ by size, cut, and oil, and I have not used a single number for them here. If you eat out, look up the chain’s published nutrition for the exact size you order, or see our restaurant guides for specific chains.

The Lund University study is the reason people sometimes say fries are better for blood sugar than boiled potatoes. In that trial of 14 healthy adults, fries gave a lower glycemic response than boiled potatoes at the same carbohydrate amount. The authors concluded that fat alone did not explain it. That does not make fries a good choice for a person with diabetes. The study did not include people with diabetes, a restaurant order is far bigger than a lab portion, and fries bring fat, calories, and sodium. The Harvard cohort data also tie higher fry intake to more weight gain than baked or mashed potatoes. If you choose fries, a small order shared, with protein and vegetables on the plate, is a more realistic plan than hoping the fat blunts the spike.

Sweet potatoes versus white potatoes

Sweet potatoes get a health halo. The carbohydrate data show why it is only half deserved. Baked sweet potato in skin has 20.7 g of carbohydrate per 100 g, nearly the same as baked russet at 21.4 g. It has more fiber, 3.3 g against 2.3 g per 100 g, and more total sugars, 6.5 g against 1.1 g. One medium sweet potato (114 g) comes to about 23.6 g of carbohydrate, and a medium boiled one (151 g) to about 26.7 g. Sweet potato’s GI of 61 in the Harvard table is lower than baked russet’s 85, but the Harvard Nutrition Source warns that its glycemic index and load are nearly as high as white potatoes, and says to keep portions small.

The sweet potato also has real strengths, including fiber and vitamin A, and the studies have not tied it to weight gain the way fries have. The best move is to treat it as a starchy food with a modest advantage, not a vegetable you can pile on.

What a portion looks like

Many people eat potatoes by size rather than grams. A rough guide: a medium baked potato is about the size of a computer mouse or a closed fist, and USDA lists it at 173 g. Half a cup of boiled potato, about 78 g, is a small scoop, roughly the size of a large egg. A cup of mashed potatoes is a heaping ice cream scoop or two. Restaurant baked potatoes are often larger than a medium potato, so a whole one can be one and a half or two servings. If you are not sure what you are eating, a food scale and one weighing session will teach your eye faster than any chart.

The ADA plate method that dietitians often teach puts starchy foods in about a quarter of the plate, with non-starchy vegetables filling half and protein a quarter. I did not re-verify the current wording of that guidance for this guide, so confirm with your care team. Under that approach, a whole large baked potato would crowd out the rest of the plate.

Pairing and order of eating

Some small studies suggest that eating vegetables and protein before the starch can soften the post-meal glucose rise. I did not pull those studies for this guide, so treat the idea as unproven for you. It is easy to test. Eat the salad, chicken, or fish first, then the potato, and compare your readings with the days you eat everything at once.

Fat, protein, and vinegar-based dressings can also slow how quickly the meal leaves your stomach, though the size of that effect varies, and I would not promise a result. A baked potato topped with butter and sour cream still has the same carbohydrate; the toppings add calories, not carbohydrate savings. A potato with beans, chili, or broccoli adds fiber and protein, which is a more useful addition than extra cheese.

A short walk after the meal is another low-cost experiment. Our guide to walking after meals covers what is known.

Potatoes if you have prediabetes

If you have prediabetes, the main question is whether potatoes are crowding out higher-fiber foods and adding to your total carbohydrate load. A smaller, boiled, or cooled portion on some days, a swap to beans or intact whole grains on others, and a consistent post-meal walk are reasonable things to discuss with a clinician. Large cohort studies tie frequent fries and large potato servings to weight gain and diabetes risk, which is a reason to watch the pattern, not a reason to ban the food. Ask your clinician or a dietitian what a sensible weekly pattern looks like for you.

Potatoes if you have type 2 diabetes and do not take insulin

For type 2 diabetes managed with food, activity, or non-insulin medicines, the approach is a carbohydrate budget. If your dietitian gave you a number per meal, a potato portion has to fit inside it. A small boiled portion can leave room for fruit or other starch. A whole baked potato may take most of the meal’s budget on its own. Pair it with non-starchy vegetables and protein, check readings one and two hours after eating if your care team recommends it, and keep notes. If you take a sulfonylurea, ask your clinician before making big changes to how much carbohydrate you eat.

Potatoes if you use a CGM

A continuous glucose monitor is the best tool for settling the potato question for yourself, because the tables disagree so widely. A fair test uses the same portion, the same time of day, and similar activity on different days. Try boiled versus baked, hot versus cooled and reheated, or potato first versus potato last. Look at the rise, the peak, and how long glucose stays elevated, and remember that sensor readings lag blood glucose by several minutes and can be less accurate during rapid changes. Our CGM guide explains the basics. If you use insulin, do not change your dosing pattern based on your own experiments without your care team, since a food that spikes you less can also cause a low if the dose assumed a bigger rise.

Worked example: one 45 g carbohydrate dinner

Suppose, purely as an example, that a dinner budget is 45 g of carbohydrate. This is not a recommendation. Your number comes from your care team. Here is what each potato choice leaves for everything else, using the USDA-based figures above.

A whole medium baked russet leaves about 8.0 g, which is less than one slice of bread, so a baked potato dinner with a roll and dessert goes over quickly. A half potato leaves about 26.5 g, enough for a small serving of fruit or beans. Twenty fries use about as much as a whole baked potato. Each of these potatoes also adds different fiber, fat, and calories, which the carbohydrate count does not show.

Common mistakes

• Treating a potato as a vegetable. Nutritionally it behaves like a starch, closer to rice or bread than to broccoli.
• Not knowing the size. A restaurant potato can be double the 173 g USDA medium.
• Assuming sweet potatoes are low-carb. They are not, at about 23.6 g for one medium.
• Counting only the potato. Gravy, buns, sauces, and sugary drinks often make up the rest of the carbohydrate in a potato meal.
• Believing one GI number. The same potato scored 50 in one table and over 100 in a study.
• Trusting the cooling trick too much. The one study I found had eight people and showed no glucose difference.
• Skipping your own data. A meter or CGM test settles it faster than the internet.

Myth versus fact

Questions to bring to your appointment

• How many carbohydrate grams should my dinner have, and where do potatoes fit?
• Should I choose boiled, baked, or cooled potatoes, and how big a portion?
• Do my medicines change how I should plan a starch-heavy meal?
• Would a CGM or a few extra finger-stick checks help me learn my response to potatoes?
• Is potassium a concern for me?

Frequently asked questions

Can I eat potatoes if I have diabetes?

Many people with diabetes eat potatoes. They are a starchy food, so what matters is the portion, how they are cooked, and what else is on the plate. A medium baked russet with the skin has about 37.0 g of carbohydrate (USDA data for a 173 g potato), which is a large share of many meal budgets. Ask your care team how many carbohydrate grams fit your meals, then check your own readings.

Are sweet potatoes better than white potatoes for blood sugar?

Not automatically. Per 100 g, baked sweet potato and baked white potato have almost the same carbohydrate (about 20.7 g vs 21.4 g in USDA data). Sweet potato has more fiber per 100 g (3.3 g vs 2.3 g) and the glycemic index tables put it somewhat lower than baked russet, but Harvard’s Nutrition Source notes that sweet potatoes still have a glycemic index and load nearly as high as white potatoes. Portion size matters most.

Is a baked potato or a boiled potato better?

In the glycemic index tables, baked russet potatoes scored higher (about 85) than average boiled white potatoes (about 50), but those values vary a lot between studies and potato varieties, and a Swedish study of healthy adults measured boiled potatoes very high. Boiled, cooled, and eaten in a smaller portion is a reasonable thing to try, then check your own response.

Do french fries raise blood sugar less than baked potatoes?

One small Swedish study of 14 healthy adults found fries gave a lower glycemic response than boiled potatoes at the same 50 g of carbohydrate. That does not make fries a good choice. A serving is easy to overeat, they add fat and salt, and the study did not include people with diabetes. Compare the carbohydrate grams, not just the glycemic index.

Does cooling a potato really lower its glycemic effect?

Cooling and reheating may form resistant starch, but the human evidence is thin. In one small crossover study of 8 people, a chilled-and-reheated mashed potato meal produced a 24% lower insulin response, but no significant difference in glucose. It is worth a test on yourself with a meter or CGM, but it is not a proven fix.

Should I eat the potato skin?

The skin holds much of the fiber, and a baked russet with skin has about 2.3 g of fiber per 100 g compared with 1.5 g for the flesh alone (USDA). The skin does not remove the carbohydrate, though. Skin from the same potato adds fiber, not a free pass on portion size.

What portion of potato is reasonable?

There is no single right amount. Many dietitians teach a portion about the size of a computer mouse or a closed fist, which is close to a medium potato, and some people split one potato across two meals. The best target depends on your carbohydrate budget, medicines, and glucose response, so ask a dietitian.

Are potatoes allowed on a low-carb diet?

Potatoes are not low-carb. A medium baked potato has about 37.0 g of carbohydrate, which could use up most or all of a strict low-carb day. If you follow a lower-carb eating pattern, a small boiled portion or a swap such as cauliflower mash may fit better.

Is it safe to eat potatoes with kidney disease or high potassium?

Potatoes are high in potassium, about 550 mg per 100 g for baked russet with skin in USDA data. People with kidney disease or on certain medicines are sometimes told to limit potassium. That is a clinician-level question, so ask your care team before changing how much you eat.

Sources

• USDA FoodData Central, SR Legacy entries (FDC IDs 170030, 170033, 170114, 170134, 168484, 170493, 169264, 170034), viewed through the getfoodfacts.com mirror
• Harvard Health glycemic index and load table (based on Foster-Powell 2002)
• International Tables of Glycemic Index and Glycemic Load Values 2021 (summary)
• Harvard T.H. Chan School of Public Health, The Nutrition Source: The problem with potatoes
• Lund University: glycaemic and satiating properties of potato products
• AGRIS record: reheated starchy meal and postprandial glycaemia (8-person crossover study)

About the author: Charles Kirkland is the founder of Metabolic Clarity Labs and author of The Journey Back. He is not a physician. This is general education, not a medically reviewed treatment plan.

Sources checked October 31, 2026. Discuss personal decisions with a qualified clinician. Do not start, stop, or change prescribed treatment based on this article.