Are Sweet Potatoes Good For Weight Loss? Honest Evidence Review

Sweet potatoes are modestly weight-loss-friendly — they are low energy density (~90 kcal per 100 g baked, ~76% water), have a lower glycemic index than white potato when boiled (~46 vs ~78 per Atkinson 2021^[1]), and deliver roughly 384% DV vitamin A per cup from beta-carotene. They are a sensible substitution tool — trading in for white potato, white rice, or pasta inside a calorie deficit. They are not a fat-burner, and there is no peer-reviewed randomized trial of any “sweet potato diet”. The Bertoia 2015 PLoS Med analysis of 133,468 US adults^[3] found total vegetables tracked with -0.25 lb of weight per +1 serving/day per 4-year period and total fruits -0.53 lb — but starchy vegetables (corn, peas, and potatoes including sweet potatoes, pooled) tracked with +1.00 lb when added on top of an existing diet, reversing only when the starchy vegetable substituted for refined grains or other starches. Magnitude vs pharmacotherapy: STEP-1 semaglutide produced -14.9% body weight at 68 weeks^[8], SURMOUNT-1 tirzepatide -20.9% at 72 weeks^[9]. A sweet potato will not produce those numbers; here is what the evidence actually supports.

The honest answer

Sweet potatoes are a sensible starchy-vegetable swap for white potato, rice, or pasta inside a calorie deficit — ~90 kcal per 100 g baked, low GI (~46 boiled), 3.3 g fiber per 100 g, and a large dose of beta-carotene. They are not a weight-loss intervention, and the “sweet potato diet” has no randomized trial evidence behind it. Portions and toppings still do all the work.

At a glance

• Baked sweet potato, flesh and skin (USDA FDC^[10]): ~90 kcal, 20.7 g carb, 3.3 g fiber, 6.5 g sugar, 2.0 g protein, 0.15 g fat per 100 g. A medium 130 g tuber is ~117 kcal. One cup (~200 g) carries roughly 384% DV vitamin A from beta-carotene.
• Sweet vs white potato GI (Atkinson 2021^[1]): boiled sweet potato GI ~46 (low-GI band) versus boiled/baked white potato GI ~78-95. The gap narrows when sweet potato is baked (~64) and disappears when it is fried or candied.
• No randomized trial of the “sweet potato diet”. The Japanese fad framing (sweet-potato-as-staple replacement) has no peer-reviewed weight-loss RCT. Calorie deficit still does the work.
• Holt 1995 satiety index excluded sweet potatoes. The famous “potato is #1 satiety” finding (boiled white potato scored 323) does not extend to sweet potato — which Holt did not test. Anyone citing a sweet-potato satiety score to Holt is inventing it.
• Cohort signal is for substitution, not addition. Bertoia 2015 PLoS Med^[3]: starchy vegetables added on top of existing intake tracked with +1.00 lb per 4-year period; the protective signal appears when starchy vegetables replace refined grains, fries, or sugary snacks.
• GLP-1 compatible. Plain boiled or mashed (without butter) sweet potato is bland, gentle on a slowed digestive tract, and the fiber + beta-carotene profile complements the protein-first GLP-1 plate.

Sweet potato nutrition per 100 g and per medium tuber

Per USDA FoodData Central^[10], baked sweet potato with skin (no added fat or salt):

• ~90 kcal per 100 g — 20.7 g carbohydrate, 3.3 g fiber, 6.5 g sugar, 2.0 g protein, 0.15 g fat, 76% water by weight.
• Per medium baked tuber (~130 g): ~117 kcal, 27 g carb, 4.3 g fiber, 8.5 g sugar, 2.6 g protein.
• Per cup (~200 g): ~180 kcal, 41 g carb, 6.6 g fiber, 4.0 g protein, ~384% DV vitamin A (19,218 IU per 100 g, essentially all as beta-carotene), 950 mg potassium.
• Boiled sweet potato, no skin: ~76 kcal per 100 g (lower than baked because no moisture loss), 17.7 g carb, 2.5 g fiber.

Two callouts that matter for a weight-loss plate. First, the sugar number (~6.5 g per 100 g baked, ~8.5 g per medium tuber) is the highest of any common starchy staple — it is what people are tasting when they call sweet potato “sweet”. This is not a problem for weight loss at typical portions; it is slow-digesting sugar inside a fibrous matrix, not added sugar. Second, the vitamin A figure is unusually high for a single food: one medium tuber covers roughly 200% of the adult DV. There is no weight-loss benefit attributed to beta-carotene specifically, but it is a real nutritional bonus relative to white potato (which has essentially zero vitamin A).

Sweet potato vs white potato — the honest comparison

Sweet potatoes are routinely marketed as the “healthier” choice over white potatoes. The truth is more nuanced.

• Calories: roughly tied. Baked white potato with skin is ~93 kcal per 100 g; baked sweet potato is ~90 kcal per 100 g (USDA FDC^[10]). The difference is inside the noise of how you weigh and prepare your tuber.
• Glycemic index: sweet potato wins when boiled. Atkinson 2021^[1] reports boiled sweet potato GI ~46 versus boiled white potato GI ~78. Baked, the sweet potato moves up to ~64 (still lower than baked white at ~85-95). For people without diabetes, this is a marginal weight-loss lever; for prediabetes or T2D, it is more meaningful.
• Fiber: sweet potato edges ahead. 3.3 g per 100 g baked vs 2.2 g for white potato — a real but modest difference.
• Protein: white potato edges ahead. 2.5 g per 100 g vs 2.0 g for sweet potato. Neither is a protein source.
• Vitamin A: sweet potato dominates. ~384% DV per cup baked from beta-carotene; white potato has essentially none.
• Satiety: probably similar. Boiled white potato is the highest-satiety food in the Holt 1995 index (score 323). Holt did not test sweet potato. Most modern satiety panels that have included sweet potato show it in the high-satiety band — high water content (76%), moderate fiber, slow glycemic profile — but a verbatim “sweet potato satiety = 323” figure does not exist in the published literature.

The practical translation: sweet vs white is a marginal optimization. If you already eat white potatoes and like them, keep eating them — see our white potato weight loss evidence review for the full satiety + cohort case. If you like sweet potatoes, eat them for the beta-carotene bonus and the lower boiled GI. Either way, the cooking method (boiled/baked vs fried/candied) and the toppings (plain vs marshmallow-and-brown-sugar) matter more than the choice of tuber.

Sweet potato vs other starchy staples

Plain baked sweet potato is the lowest-calorie option in this comparison set — tied with white potato at ~90 kcal per 100 g. Quinoa carries more complete protein per gram (see our quinoa for weight loss evidence review); beans dominate fiber and protein (beans evidence review). Pasta is the highest-calorie option per 100 g cooked. The ranking flips entirely once preparation is taken into account: sweet potato fries land in the same calorie-density range as pasta, and candied sweet potatoes can reach 200+ kcal per 100 g.

The “sweet potato diet” fad — what the evidence shows

Variations of a “sweet potato diet” circulate in popular nutrition writing — the Japanese imojin framing in which sweet potato substitutes for rice as the staple carbohydrate, the “3-day sweet potato cleanse”, and viral social-media protocols using sweet potato as a near-monodiet. None of these has a peer-reviewed randomized weight-loss trial in humans. A PubMed search for “sweet potato” combined with “weight loss” or “body weight” in randomized trials returns small studies of sweet-potato extract components (caiapo / Ipomoea batatas extract for glycemic control), animal models, and observational nutrient-pattern analyses — no human RCT testing whole sweet potato as a weight-loss intervention.

What the cohort literature does support is more modest: substituting starchy vegetables (sweet potato included) for refined grains or fried side dishes is associated with small favorable weight-trajectory shifts. The Bertoia 2015 PLoS Med analysis^[3] of 133,468 US adults found total vegetables tracked with -0.25 lb per +1 serving/day per 4-year period and total fruits -0.53 lb — but starchy vegetables (the pooled category containing corn, peas, and potatoes including sweet potatoes) tracked with +1.00 lb when intake increased on top of existing diet. The protective signal appears only when the starchy vegetable substitutes for a denser food, not when it adds calories. That is the core substitution math any “sweet potato diet” would depend on, and it is small.

The Schlesinger 2019 Adv Nutr meta-analysis^[5] of prospective studies on food groups and weight gain placed vegetables (which include starchy tubers as part of the broader category) at RR 0.91 for overweight/obesity per 100 g/day increment — small but favorable. The Aune 2017 Int J Epidemiol meta^[4] of 95 prospective studies found each 200 g/day of fruits + vegetables reduced all-cause mortality RR 0.90; sweet potato counted as vegetable but was not separately analyzed. The pattern is real: more vegetables, a small favorable signal. The pattern is not specific to sweet potato.

Mediterranean and DASH integration

Sweet potato fits cleanly into both Mediterranean and DASH eating patterns as a starchy-vegetable side. The Estruch 2018 N Engl J Med PREDIMED reanalysis^[7] of 7,447 high-CV-risk adults reported a 28% reduction in major cardiovascular events (HR 0.72) in the Mediterranean arms vs control over a median 4.8 years, with neutral weight outcomes. The Mediterranean pattern emphasizes vegetables (including starchy tubers), legumes, whole grains, fish, olive oil, and nuts — with sweet potato landing as a reasonable starchy side. The Bes-Rastrollo 2007 SUN cohort^[6] of 8,865 Mediterranean-pattern adults showed frequent nut consumers gained 0.78 kg less weight over 28 months; a similar pattern of modest favorable weight signal applies to whole-vegetable intake in adjacent SUN analyses.

DASH (Dietary Approaches to Stop Hypertension) emphasizes potassium-rich vegetables and fruits; sweet potato is excellent on this axis (~475 mg potassium per 100 g baked, ~950 mg per cup). For people focused on blood-pressure control, sweet potato is a useful side. For weight loss specifically, the dietary-pattern signal is small — single-digit kilograms over multi-year horizons.

Pairing sweet potato with GLP-1 therapy

For patients on semaglutide, tirzepatide, or other GLP-1 receptor agonists, sweet potato is one of the better-tolerated carbohydrate sides. Practical guidance:

• Plain boiled or mashed (no butter) sweet potato is bland, low-fat, and gentle on a slowed digestive tract. The fiber load (~3 g per medium tuber) is moderate — enough to help with GLP-1-induced constipation but not so much that it triggers bloating.
• One medium tuber per meal (~117 kcal plain, ~150 kcal with a teaspoon of olive oil) sits well alongside a protein anchor (chicken, salmon, eggs, Greek yogurt, lentils). See our what to eat on a GLP-1 protein guide for plate structure.
• Skip the fries. Sweet potato fries trigger GLP-1 nausea at higher rates than baked sweet potato — the oil content slows gastric emptying further on a drug that is already slowing it. They also undercut the calorie math.
• Avoid candied / marshmallow preparations. Thanksgiving-style sweet potato casserole (200-350 kcal per cup) is a dessert dressed as a side. It is poorly tolerated on GLP-1s and is calorically a different food than plain sweet potato.
• Beta-carotene + fiber bonus. GLP-1 patients often report difficulty getting enough vegetables in once appetite is suppressed; sweet potato is calorie-dense enough to fit a small portion in even on a 1,200-1,500 kcal day while delivering vitamin A.

For broader GLP-1 side-effect troubleshooting, see our GLP-1 side effects Q&A hub. Plain sweet potato shows up frequently in the well-tolerated-foods list for nausea-prone patients.

When sweet potatoes backfire

Three preparation categories convert sweet potato from a modest weight-loss-friendly side into a calorie problem:

• Sweet potato fries. ~250 kcal per 100 g at most US restaurants — roughly 2.8x the calorie density of baked. A typical restaurant side portion (130-180 g) lands at 325-450 kcal before the meal's entree. The “sweet potato fries are healthier than regular fries” framing is largely marketing; deep-fried oil-soaked starch is roughly the same calorie-density problem in either tuber.
• Candied / marshmallow / brown-sugar preparations. The Thanksgiving casserole format (~250-350 kcal per cup) buries the tuber under butter, brown sugar, and marshmallow topping. The added sugar load (15-25 g per serving) flips the food from low-GI staple to a dessert calorie-equivalent.
• Oversized portions. Restaurant-served sweet potato is routinely 200-300 g (a 1.5-2x typical home portion), often arriving split open with butter pre-melted. A reasonable home portion is one medium tuber (130 g, ~117 kcal); a steakhouse portion can land at 250+ kcal before toppings.

Magnitude vs GLP-1 pharmacotherapy

For context: the Wilding 2021 STEP-1 trial of semaglutide 2.4 mg weekly^[8] reported -14.9% body weight at 68 weeks. The Jastreboff 2022 SURMOUNT-1 trial of tirzepatide 15 mg weekly^[9] reported -20.9% at 72 weeks. For a 100-kg starting weight, those are -15 kg and -21 kg. No whole-food intervention — sweet potato or otherwise — has produced anything in that magnitude range in a randomized trial. The sweet-potato question lives at the substitution-math layer (replacing fries with baked sweet potato, or replacing white rice with mashed sweet potato in a bowl) — small, real, single-digit pounds over years.

Verdict and bottom line

• Yes, sweet potatoes are good for weight loss — modestly. ~90 kcal per 100 g baked, low GI (~46 boiled) per Atkinson 2021^[1], 3.3 g fiber, ~384% DV vitamin A per cup. A sensible starchy-vegetable substitute for white potato, rice, or pasta inside a calorie deficit.
• Better than white potato on three axes (GI, fiber, vitamin A) and roughly tied on calories. The sweet-vs-white question is a marginal optimization; both are reasonable.
• No randomized trial of the “sweet potato diet”. The Japanese fad framing has no peer-reviewed human weight-loss RCT evidence behind it. Calorie deficit still does the work.
• Preparation is the lever, not the tuber. Baked or boiled with skin: ~90 kcal per 100 g. Fried: ~250 kcal per 100 g. Candied with marshmallow: 200-350 kcal per cup. Same food, different calorie tier.
• GLP-1 compatible. Plain boiled or mashed sweet potato is one of the better-tolerated carbohydrate sides for patients on semaglutide or tirzepatide. One medium tuber per meal alongside a protein anchor is a reasonable plate.
• Magnitude check. No food — sweet potato included — approaches the -14.9% to -20.9% body weight reduction of semaglutide^[8] or tirzepatide^[9]. Sweet potato is a food, not a treatment.

Related research and tools

• Are potatoes good for weight loss? (white potato evidence review) — the sibling article on white potato, with the Holt 1995 satiety story (boiled white potato scored 323, highest of 38 foods tested) and the Mozaffarian 2011 fries-vs-baked gap.
• Is quinoa good for weight loss? — the complete-protein pseudocereal alternative to a sweet potato side.
• Are beans good for weight loss? — the high-fiber, high-protein legume option that pairs naturally with sweet potato in a bowl.
• Is peanut butter good for weight loss? — companion article on the nut/seed protein-fat anchor for a starch-plus-protein plate.
• GLP-1 side effects Q&A hub — tolerability framing for sweet potato as a bland carbohydrate during nausea-heavy weeks.
• What to eat on a GLP-1: protein-first guide — meal-pattern structure sweet potato fits as the starch side.

Important disclaimer. This article is educational and does not constitute medical or nutrition advice. Patients with type 2 diabetes, prediabetes, or hypertension should discuss carbohydrate choices, glycemic index considerations, and portion sizes with their clinician or a registered dietitian. Patients on GLP-1 therapy who experience persistent nausea, vomiting, or early satiety should not push through with sweet potato or any other food — contact the prescribing clinician. PMIDs were independently verified against the PubMed E-utilities API on 2026-05-19; USDA per-100-g values were taken from FoodData Central entries for baked and boiled sweet potato and reflect general supermarket produce. Variety (orange-flesh vs purple-flesh vs Japanese white-flesh), growing conditions, and preparation can shift these numbers materially.

Last verified: 2026-05-19. Next review: every 12 months, or sooner if new RCT evidence on sweet potato intake and weight outcomes is published.