Converting BMI Units: How Metric, Imperial, and Stones Rounding Changes a Category

A rounding error of 0.2 kg/m² can push a person from the WHO normal range (18.5-24.9) into the overweight band (25.0-29.9) without any change in their actual body. Converting BMI between metric, imperial, and stones introduces this risk at every step, because you are converting weight and height between systems that use different base units and different rounding conventions. The conversion factors are fixed: 1 kg = 2.20462 lb, 1 stone = 14 lb, 1 inch = 0.0254 m. Use these exact numbers, not approximations. Round only at the final BMI value to one decimal place. Rounding any intermediate number, weight in pounds, height in inches, stones to the nearest quarter-stone, introduces a systematic shift of 0.1-0.3 kg/m². That shift is large enough to cross a WHO boundary.

The BMI formula is weight in kilograms divided by height in metres squared. Quetelet devised this index in the 19th century as a statistical descriptor, not a health metric. The WHO adult BMI cut-offs from 1995, underweight below 18.5, normal 18.5 to 24.9, overweight 25 to 29.9, obese 30 and above, were set by an expert committee reviewing epidemiological studies from Europe and North America. The arithmetic itself does not change. What changes is the precision of your inputs.

You get a BMI in kg/m². That is the only unit BMI uses. Every conversion from imperial or stones must produce a weight in kilograms and a height in metres before you can apply the formula. The error enters when you round the converted numbers before dividing.

The Exact Conversion Factor

One kilogram equals 2.20462 pounds. To convert pounds to kilograms, divide the weight in pounds by 2.20462. Do not use 2.2. The difference between 2.20462 and 2.2 is 0.21% per conversion, which sounds trivial but compounds when you then divide by height squared. A person weighing 180 lb gives 81.65 kg using the full factor and 81.82 kg using 2.2, a difference of 0.17 kg.

What That Does to BMI

A man who is 180 lb and 70 inches tall (1.778 m) has a BMI of 25.82 using the full conversion. Using 2.2, his BMI lands at 25.88. Both fall in the overweight band, so no harm done here. But a woman weighing 165 lb (74.84 kg) at 64 inches (1.626 m) has a BMI of 28.29. Using 2.2, her weight becomes 75.0 kg and her BMI 28.38. Still overweight, no boundary crossed. The risk appears at the boundary between normal and overweight. A person at BMI 24.8 using full conversion can hit 25.1 using 2.2, crossing into overweight on a rounding error.

Failure Case

The reader who rounds pounds to the nearest half-pound before converting introduces a second error. Self-reported weight already carries a systematic downward bias of roughly 0.5-1.5 kg/m² because people under-report weight and over-report height. Adding a rounding convention on top of that bias pushes the final BMI further from the truth. The correct method: convert the exact self-reported weight in pounds using 2.20462, keep three decimal places through the calculation, and round only the final BMI to one decimal place.

Stones and pounds add another layer. One stone is exactly 14 lb. A person who says they weigh 12 stone 7 lb is giving a weight in half-stones. The common rounding to the nearest quarter-stone, 12 stone 6 lb or 12 stone 8 lb, introduces an error of up to 0.5 kg before you even convert to kilograms.

Worked Example in All Three Systems

Take a person whose measured weight is 12 stone 5 lb (173 lb) and measured height is 5 foot 7 inches (67 inches). First, convert to kilograms: 173 lb ÷ 2.20462 = 78.47 kg. Convert height to metres: 67 in × 0.0254 = 1.7018 m. Square the height: 1.7018² = 2.8961. Divide: 78.47 ÷ 2.8961 = 27.10 kg/m². That sits in the overweight band.

Now round the weight at the stones-and-pounds step. If the person reports 12 stone 6 lb (174 lb), the weight becomes 78.93 kg and BMI becomes 27.27. Still overweight, no category change. If the person reports 12 stone 4 lb (172 lb), weight becomes 78.01 kg, BMI becomes 26.94. Still overweight.

The boundary-crossing risk appears at the top of the normal range. A person at 10 stone 12 lb (152 lb) and 5 foot 7 inches has a weight of 68.95 kg and a BMI of 23.81. If they round up to 11 stone even (154 lb), weight becomes 69.85 kg and BMI becomes 24.12. Still normal, safe. But at 65 inches height (1.651 m), the same 10 stone 12 lb gives BMI 25.30. Rounding to 11 stone gives 25.63, both already overweight. The rounding did not change the category, but the arithmetic shows how close the boundary is.

Where It Breaks

The real failure mode is the person who reports their weight as "about 11 stone" without the pounds. That introduces an error of up to 6.5 lb (2.95 kg), which can shift BMI by 1.0-1.2 kg/m², enough to move from the middle of the normal range to the middle of overweight. Do not accept a weight stated only in whole stones. Ask for the pounds or, better, get the weight in kilograms from a calibrated scale.

Height is not stable. Diurnal height variation causes a loss of up to 1.5 cm from morning to evening due to spinal disc compression. A person measured at 180 cm in the morning may be 178.5 cm by evening. That 1.5 cm difference changes BMI by roughly 0.5 kg/m², the same magnitude as the self-report bias.

One inch equals exactly 0.0254 m. To convert feet and inches to metres, multiply the total inches by 0.0254. A person who is 5 foot 10 inches (70 inches) becomes 1.778 m. A person who says they are "about 5 foot 10" may actually be 69.5 inches or 70.5 inches, an error of 0.5 inches (0.0127 m). That alone changes BMI by about 0.15 kg/m². Combined with weight rounding, the total error reaches 0.3-0.5 kg/m².

Worked Example

Take a person with measured height 69.5 inches (1.7653 m) and measured weight 155 lb (70.31 kg). BMI: 70.31 ÷ (1.7653²) = 70.31 ÷ 3.1162 = 22.56. Normal range. If the person self-reports as 70 inches and 155 lb, height becomes 1.778 m, weight is the same 70.31 kg, and BMI becomes 70.31 ÷ 3.1608 = 22.24. Still normal, the 0.32 kg/m² difference does not cross a boundary. But if the person is actually 65.5 inches and 175 lb, BMI is 28.57. Self-reporting as 66 inches and 170 lb drops BMI to 27.62, still overweight, but the error is large enough to hide a borderline class I obesity (BMI 30.0) if the true weight were higher.

The WHO adult BMI cut-offs have sharp boundaries: 24.9 is normal, 25.0 is overweight. A BMI of 24.95 rounds to 24.9 if you round to one decimal place and to 25.0 if you round to the nearest whole number. The convention in clinical and population research is to report BMI to one decimal place and to use the unrounded value for classification. The World Health Organization Technical Report Series 854 specifies the bands as inclusive of the lower bound and exclusive of the upper: 18.5 is included in normal, 25.0 is included in overweight. That means a BMI of exactly 25.0 is overweight, not normal.

Rounding at intermediate steps, as shown in the worked examples, adds 0.1-0.3 kg/m² of error. The most dangerous combination is rounding weight down (self-report bias) and rounding height up (self-report bias plus diurnal variation). A person whose true BMI is 24.9 can end up with a calculated BMI of 25.2, crossing into overweight. Conversely, a person at 25.1 can end up at 24.8, appearing normal when they are overweight. Both misclassifications happen routinely in self-reported surveys and in health apps that accept rounded inputs.

For the Person With a BMI Number

You have a BMI number from a doctor, an app, or a scale and you want to know what it actually means without being sold a programme. This reference tells you which organisation set which threshold and when, so you can judge how much weight to put on the number.

For Fitness and Health Professionals

You need the provenance of every threshold to explain to clients without overclaiming. Every cut-off here is traced to its source document and year.

For the Data-Literate Reader

You have seen BMI criticised online and you want to know which criticisms are supported by meta-analyses and which are category errors. This reference separates the arithmetic from the interpretation.

For Writers and Editors

You are building content about body measurement and you need a single source of truth for which body published which cut-off in which year.

This reference does not suit someone looking for a diet plan, a calorie target, or a workout schedule. It does not suit someone trying to decide whether to take a weight-loss drug. It does not suit someone who wants to know if they are healthy based on BMI alone, the site explains why that question cannot be answered by a single anthropometric number. The single thing that most often goes wrong here is treating a BMI number as a direct readout of metabolic health or body fatness for a specific person, when it was designed to spot statistical shifts in large groups.