How to Build a Nutrient-Dense Plate Using Raw USDA Data
Use actual per-calorie nutrient data instead of guesswork to decide what goes on your plate β here's the method, the real numbers, and the tool that applies it for you.
A nutrient-dense plate isn't about eating "healthy-sounding" foods β it's about maximizing nutrients per calorie, not per gram. USDA data shows lettuce delivers nearly 1,500 mg of potassium per 100 kcal, far more than most other vegetables calorie-for-calorie. This guide explains the method, then points you to FoodData Explorer's own calculator to apply it.
What does "nutrient-dense" actually mean β and why isn't it about weight?
"Nutrient-dense" gets used loosely enough that it's worth defining precisely: a food is nutrient-dense when it delivers a lot of a given vitamin, mineral, or other nutrient relative to the calories it costs β not relative to its weight. Comparing foods per 100 grams, the way most nutrition labels and food databases default to, quietly bakes in a flaw: a 100-gram scoop of oil and a 100-gram serving of spinach aren't remotely comparable in calories, so a per-weight comparison can make a calorie-dense food look deceptively competitive with a genuinely nutrient-dense one. FoodData Explorer's own nutrient density calculator was built specifically to fix this, ranking foods by nutrient amount per 100 kilocalories instead of per 100 grams β which is the method this entire guide walks through before sending you to that tool to apply it to any nutrient you're targeting.
Why does a per-100g comparison distort the picture so badly?
Take two real foods from FoodData Explorer's own database: kale, raw carries 35 kcal per 100 g, while a calorie-dense food can carry ten times that in the same weight. If you're comparing "how much potassium am I getting," the honest question isn't "per 100 grams of each" β it's "per 100 calories of each," because calories are the actual budget most people are managing, whether for weight, energy needs, or simply fitting more actual nutrition into a fixed number of meals a day. Re-ranking by calories instead of weight is a small arithmetic change with a large practical effect: it's the difference between a ranked list that rewards bulk and one that rewards nutritional efficiency.
Which foods actually deliver the most vitamin C per calorie?
Running FoodData Explorer's own measured data for vitamin C per 100 kcal turns up broccoli, raw at roughly 294.5 mg per 100 kcal (91.3 mg of vitamin C in just 31 kcal), kale, raw close behind at about 266.9 mg per 100 kcal (93.4 mg in 35 kcal), and strawberries, raw at roughly 180.7 mg per 100 kcal. All three numbers come directly from this project's own food_nutrient and food tables, not an external estimate β and all three vegetables and fruits beat a typical calorie-dense food by a wide margin on this specific nutrient, which is exactly the comparison a per-100g ranking would obscure.
The method change doesn't just add numbers β it can flip the actual ranking order, which is the clearest proof this isn't a cosmetic distinction. Ranked by raw vitamin C per 100 grams, kale (93.4 mg) edges out broccoli (91.3 mg), with kiwifruit (74.7 mg) ahead of strawberries (56.0 mg). Ranked by vitamin C per 100 kcal instead, broccoli jumps ahead of kale (because broccoli carries slightly fewer calories per 100 g, at 31 kcal versus kale's 35), and strawberries leapfrog kiwifruit entirely (31 kcal versus kiwifruit's 58 kcal for a similar vitamin C return). Same four foods, same underlying USDA measurements β a different question about the denominator produces a different winner.
Which foods deliver the most potassium and iron per calorie?
For potassium, lettuce, cos or romaine, raw tops FoodData Explorer's own per-100-kcal ranking at roughly 1,488 mg per 100 kcal (253 mg of potassium in just 17 kcal) β ahead of kale, raw at about 994 mg per 100 kcal and broccoli, raw at about 977 mg per 100 kcal. The same romaine lettuce also leads the iron per-100-kcal ranking in this database at roughly 5.6 mg per 100 kcal, ahead of kale, raw at about 4.6 mg per 100 kcal. Low-calorie leafy vegetables repeatedly top these calorie-adjusted rankings for a straightforward reason: when the denominator (calories) is small and the nutrient amount is still meaningful, the per-100-kcal figure gets large β which is the entire mechanism this method is built to surface.
Does nutrient density mean protein-rich foods automatically "win" too?
Not automatically, and this is where the per-calorie method earns its keep for macronutrients as well as micronutrients: ranking protein per 100 kcal in FoodData Explorer's data puts lean, low-fat seafood at the top β haddock, raw delivers about 22.0 g of protein per 100 kcal (16.3 g of protein in just 74 kcal), narrowly ahead of raw pollock at about 22.0 g per 100 kcal and canned tuna in water at about 21.1 g per 100 kcal. A fattier protein source can carry a similar absolute amount of protein per 100 grams yet rank lower on this calorie-adjusted list, simply because its extra fat content raises the calorie denominator without adding more protein β again, exactly the distortion a per-weight comparison would hide.
How does this connect to Harvard's Healthy Eating Plate?
The per-calorie framing isn't just a database trick β it lines up with how Harvard T.H. Chan School of Public Health's Nutrition Source frames plate composition more broadly. Its Healthy Eating Plate model allocates about half the plate to vegetables and fruits, a quarter to whole grains, and a quarter to protein, explicitly advising readers to "aim for color and variety" in the vegetable and fruit portion rather than defaulting to the same few choices. That advice holds up well against the actual per-100-kcal data above: romaine lettuce, kale, and broccoli all land near the top of multiple nutrient rankings specifically because they're low-calorie foods carrying a meaningful absolute nutrient amount, which is the same property that makes "half your plate" in vegetables and fruits a nutrient-dense allocation rather than an arbitrary one.
So how do you actually build a plate using this method?
Pick the nutrients you're actually trying to increase β potassium, iron, vitamin C, protein, or any of the 100-plus nutrients and compounds tracked in this database β and run each one through FoodData Explorer's nutrient density calculator rather than guessing from a food's reputation. The tool exists precisely so you don't have to re-run the kind of calorie-adjusted ranking query this guide just walked through by hand: pick a nutrient from its dropdown, and it returns the foods that deliver the most of it per 100 kcal, using the same USDA FoodData Central Foundation Foods data (release 2026-04-30) cited throughout this guide.
A worked example pulls the method together: someone building a lower-calorie plate that still covers potassium, iron, and protein could pair a base of romaine lettuce or kale β both top-ranked for potassium and iron per 100 kcal in the data above β with a lean protein source like haddock, which topped the per-100-kcal protein ranking without the added calorie cost of a fattier cut. Running each nutrient separately through the calculator before building the plate, rather than assuming a food is "good for" a nutrient because it has a healthy reputation, is the actual behavior change this guide is arguing for. The method matters more than memorizing any single ranking, because the "best" food for your plate depends entirely on which nutrient you're short on β a question only you and, where a documented health condition is involved, your clinician can answer; this guide's job is to make sure the comparison itself is done honestly, per calorie, once you know what to look for.
Related database pages
Frequently asked questions
- What's the actual difference between "nutrient-dense" and just "low-calorie"?
- A low-calorie food isn't automatically nutrient-dense; it only earns that label if it also delivers a meaningful amount of a given nutrient. Nutrient density specifically measures nutrient amount per 100 kilocalories, so a low-calorie food with very little of a given nutrient would still rank poorly on that specific nutrient's list.
- Why do leafy greens like lettuce and kale show up at the top of so many nutrient-density rankings?
- Leafy greens combine a genuinely low calorie count with meaningful absolute amounts of several vitamins and minerals, and dividing a meaningful numerator by a small denominator produces a large per-100-kcal result. That's a mathematical consequence of the ranking method, not a claim that leafy greens are the single best source of every nutrient.
- Does a food with more protein per 100 grams always have more protein per 100 calories too?
- Not necessarily. A fattier protein source can match a leaner one in grams of protein per 100 grams while still ranking lower per 100 kilocalories, because the extra fat adds calories without adding protein, which lowers its calorie-adjusted ranking.
- Can I use the nutrient density calculator to find the best food for any nutrient, not just the ones in this guide?
- Yes. The calculator covers more than 100 nutrients and compounds in FoodData Explorer's database, not just the four highlighted in this guide, so it can be used for any nutrient you're specifically trying to increase.
- Should I only eat the top-ranked foods from a nutrient density search?
- No single ranking should dictate an entire diet. The ranking is a tool for finding efficient sources of a specific nutrient you're targeting, not a complete eating plan on its own, and overall variety and total calorie needs still matter.
Written by Phuoc Huynh Β· Reviewed by Phuoc Huynh on 2026-10-04
Last verified against USDA FoodData Central β Foundation Foods.
USDA release 2026-04-30. All values are per 100Β g unless stated otherwise.