LDL cholesterol is calculated, not directly measured, on most standard lipid panels. The classic formula is simple: take your total cholesterol, subtract your HDL cholesterol, and subtract your triglycerides divided by 5. All values need to be in mg/dL. If your total cholesterol is 200, your HDL is 50, and your triglycerides are 150, your LDL would be 200 − 50 − (150 ÷ 5) = 120 mg/dL.
The Friedewald Equation
For decades, the standard method has been the Friedewald equation:
LDL = Total Cholesterol − HDL − (Triglycerides ÷ 5)
The “triglycerides divided by 5” part is an estimate of your VLDL cholesterol, another type of cholesterol-carrying particle in your blood. The formula essentially says: take all the cholesterol in your blood, subtract the “good” cholesterol (HDL), subtract the VLDL estimate, and what’s left is your LDL.
This works well for most people, but it has real limits. The formula assumes a fixed relationship between triglycerides and VLDL cholesterol, which isn’t true for everyone. It becomes unreliable when triglycerides are above 400 mg/dL, and it also loses accuracy at very low LDL levels (below about 70 mg/dL). It requires a fasting blood sample, since eating raises triglycerides and throws off the estimate.
Newer, More Accurate Formulas
The 2026 ACC/AHA cholesterol management guidelines now recommend two newer equations over Friedewald for estimating LDL from a standard lipid panel: the Martin-Hopkins equation and the Sampson-NIH equation.
The Martin-Hopkins method uses the same basic lipid panel numbers but replaces the fixed “divide by 5” with a personalized adjustment factor. Instead of assuming everyone’s triglyceride-to-VLDL ratio is 5:1, it pulls from a table of 174 different ratios based on your specific triglyceride and non-HDL cholesterol levels. This makes it significantly more accurate for people with high triglycerides or very low LDL. In studies comparing fasting and nonfasting samples, this approach maintained 87% to 94% accuracy across all LDL categories, while the Friedewald method dropped to 71% accuracy for patients with LDL below 70 mg/dL.
The Sampson-NIH equation, developed at the National Institutes of Health, extends accurate LDL calculation to patients with triglycerides up to 800 mg/dL, double the Friedewald limit. It reduced misclassification of patients into the wrong treatment category by 35% when triglycerides were between 400 and 800 mg/dL.
You can’t easily do either of these calculations by hand. Many labs and online calculators now offer them automatically. If your lab report still uses the Friedewald method and your triglycerides are elevated or your LDL is very low, it’s worth plugging your numbers into a Martin-Hopkins calculator (Johns Hopkins hosts one freely online).
When Direct Measurement Is Better
Sometimes calculation isn’t good enough, and a direct LDL measurement is the better option. The American College of Endocrinology recommends direct measurement for people with triglycerides at or above 250 mg/dL (approximately 2.8 mmol/L), diabetes, or known vascular disease. The AHA/ACC guidelines also suggest direct measurement or a newer equation when calculated LDL falls below about 70 mg/dL, since Friedewald tends to underestimate LDL at that range.
Direct LDL testing uses chemical methods to measure LDL particles in your blood sample rather than estimating them from other numbers. It costs more and isn’t part of a routine lipid panel, so your doctor would need to order it specifically.
Fasting vs. Nonfasting Samples
The Friedewald equation was designed for fasting blood draws, typically after 9 to 12 hours without eating. Eating raises your triglyceride levels temporarily, and since triglycerides are a key input in the formula, a nonfasting sample can produce an inaccurate LDL result, particularly an underestimate.
The Martin-Hopkins method handles nonfasting samples much better. Its nonfasting accuracy for patients with LDL below 70 mg/dL was 92%, compared to 71% with Friedewald. If you had a nonfasting lipid panel drawn and your LDL was calculated using Friedewald, the number on your report may be less reliable than you’d expect.
Converting Between Units
Labs in the United States report cholesterol in mg/dL, while most other countries use mmol/L. If you need to convert your numbers before running a calculation, the conversion factors are:
- Total, HDL, and LDL cholesterol: multiply mmol/L by 38.67 to get mg/dL (or divide mg/dL by 38.67 to get mmol/L)
- Triglycerides: multiply mmol/L by 88.57 to get mg/dL (or divide mg/dL by 88.57 to get mmol/L)
Note that cholesterol and triglycerides use different conversion factors. If your results are in mmol/L and you want to use the Friedewald formula, convert all values to mg/dL first, then apply the equation.
Non-HDL Cholesterol: A Simpler Alternative
There’s an even easier number you can calculate from your lipid panel that some research suggests is a better predictor of heart disease risk than LDL alone:
Non-HDL Cholesterol = Total Cholesterol − HDL
This captures all the cholesterol carried by particles that contribute to artery plaque buildup, not just LDL. It’s useful because it also accounts for VLDL and other remnant particles. People with normal LDL but elevated non-HDL cholesterol often have a higher number of small, dense LDL particles, which carry additional risk that a standard LDL number can miss.
Non-HDL cholesterol doesn’t require fasting, doesn’t depend on triglyceride levels, and costs nothing extra to calculate. If your total cholesterol is 200 and your HDL is 50, your non-HDL cholesterol is 150 mg/dL. Guidelines generally recommend keeping non-HDL cholesterol about 30 mg/dL above your LDL target.