The short answer
A plain home-made wheat roti of about 30–35 g of dough is usually somewhere between 70 and 110 calories.
If you want one number to write down, 90 is a reasonable middle. But the range is the honest answer, and understanding why it is a range will do more for your tracking than memorising any single figure.
What actually changes the number
Size
This is the big one, and it is the variable most charts quietly ignore. A roti made in one house is not the same size as a roti made in another. Diameter varies, thickness varies more, and thickness matters disproportionately because it scales with volume rather than area.
A small, thin phulka might be 60 calories. A thick, generous roti from the same flour can be 130. That is more than a twofold difference between two things both correctly called "one roti".
Ghee or oil
A plain roti straight off the tawa has no added fat. A roti with ghee smeared on it has another 40–45 calories per teaspoon.
This one detail is responsible for an enormous amount of tracking error. Three rotis with ghee is roughly 130 calories more than three without — every single day. Over a month, at otherwise identical intake, that is the difference between a deficit that works and one that does not.
Flour
Whole wheat atta is the default. Multigrain sits slightly higher, being generally a little denser. Maida is a touch lower in fibre and marginally different in calories, but the difference is small compared with size and fat.
How it is cooked
A dry tawa roti and a paratha are different foods. Once the dough is being cooked in fat rather than on a dry surface, you are adding 50–100 calories per piece depending on how generous the pan was.
A rough table
| Item | Approximate calories |
|---|---|
| Small phulka, no ghee | 60–75 |
| Standard roti, no ghee | 80–110 |
| Standard roti with ghee | 120–155 |
| Thick roti / bhakri | 120–160 |
| Plain paratha | 150–250 |
| Aloo paratha with butter | 250–350 |
These are estimates. Anyone presenting them to the decimal place is guessing with more confidence than the evidence supports.
Why food databases handle this badly
Search "roti" in a typical calorie tracking app and you get a list of entries with confident, specific numbers, submitted by strangers whose rotis you have never seen. One says 71. One says 104. One says 297, which is somebody's paratha entered under the wrong name.
You now have to pick. You have no basis on which to pick, so you pick the middle one, and you carry that arbitrary choice for months.
The problem is not that the database lacks the food. It is that the database cannot know your kitchen, and it asks you to pretend otherwise.
The practical approach
Pick a number and stay consistent. Consistency matters far more than accuracy here. If your rotis are always roughly the same and you always count them at 95, then a change in your weight trend tells you something real, even if 95 was slightly off.
Count the ghee separately. It is the single largest correctable error, it is easy to observe, and it is the one most people skip.
Correct against your own trend. After two or three weeks of consistent logging, your weight trend tells you whether your assumptions are roughly right. That feedback is worth more than any chart.
How BetterCal handles it
You write "three rotis with ghee" and it is logged as three rotis with the ghee counted — estimated against how rotis are actually made and served where you are, rather than against a US food database average.
If your rotis are bigger or smaller than the estimate assumes, correct it once. The correction is remembered, so from then on your rotis are your rotis.
That is the useful bit. Not a perfect number on day one, but a number that gets closer to your actual kitchen the more you use it — without you searching a database and choosing between four strangers' guesses.
Related: calorie counts for other foods · an Indian food calorie chart · how BetterCal tracks Indian food · are AI calorie counters accurate?
