Light's criteria: a pleural effusion is an exudate if any one of these is true — fluid/serum protein ratio above 0.5, fluid/serum LDH ratio above 0.6, or fluid LDH above two-thirds of the upper limit of normal for serum LDH. If none is true, it is a transudate. In diuresed heart failure, a serum–effusion albumin gradient above 1.2 g/dL points back to a transudate.
This Light's criteria calculator checks all three criteria, shows how far each value is from its cutoff, flags borderline results and adds the protein and albumin gradients used when diuretics may have concentrated the fluid.
How to use this calculator
- Enter pleural fluid and serum total protein, choosing g/dL or g/L.
- Enter pleural fluid and serum LDH in U/L.
- Enter the upper limit of normal for serum LDH from your laboratory's reference range.
- If the patient has been on diuretics, add serum and pleural fluid albumin.
- Read the classification, then check the per-criterion results and any borderline warning.
The three criteria
| Criterion | Exudate if | Sensitivity | Specificity |
|---|---|---|---|
| Pleural fluid protein ÷ serum protein | > 0.5 | 91% | 89% |
| Pleural fluid LDH ÷ serum LDH | > 0.6 | 93% | 82% |
| Pleural fluid LDH | > ⅔ of serum LDH upper limit of normal | 66% | 100% |
| Light's criteria (any one) | — | 98% | 77% |
Thresholds from Light and colleagues (Annals of Internal Medicine, 1972), with the LDH criterion expressed as two-thirds of the upper limit of normal as in Light's 2002 NEJM review. Accuracy figures are from the MSD Manual table adapted from that review. The original paper used a fixed 200 IU for fluid LDH, which was two-thirds of the upper limit of normal in its laboratory; expressing it relative to your own lab avoids errors when assays differ.
Worked examples
Five sample effusions run through the calculator's own functions, with an LDH upper limit of normal of 222 U/L.
| Case | Protein ratio | LDH ratio | Fluid LDH vs 148 | Light's | Albumin gradient |
|---|---|---|---|---|---|
| Parapneumonic effusion | 0.66 | 2.38 | 620 | Exudate | 0.8 g/dL |
| Heart failure, no diuretic | 0.29 | 0.45 | 90 | Transudate | 2.4 g/dL |
| Heart failure after diuresis | 0.54 | 0.64 | 135 | Exudate (borderline) | 1.7 g/dL |
| Malignant effusion | 0.68 | 1.78 | 410 | Exudate | 0.6 g/dL |
| Cirrhosis (hepatic hydrothorax) | 0.23 | 0.37 | 70 | Transudate | 1.9 g/dL |
The diuresed heart failure case shows the classic trap. Its protein ratio is 0.54 and its LDH ratio 0.64, both just over their cutoffs, so Light's criteria call it an exudate — but nothing is more than 15% past its cutoff, so the calculator marks it borderline. The protein gradient is 3.2 g/dL (above 3.1) and the albumin gradient is 1.7 g/dL (above 1.2), and both point back to a transudate.
Common causes of each type
| Transudates | Exudates |
|---|---|
| Heart failure | Pneumonia (parapneumonic effusion, empyema) |
| Cirrhosis (hepatic hydrothorax) | Cancer (lung, breast, lymphoma, mesothelioma) |
| Nephrotic syndrome | Tuberculosis |
| Peritoneal dialysis | Pulmonary embolism (often exudate) |
| Hypoalbuminaemia | Autoimmune disease, pancreatitis |
Heart failure is the most common cause of a transudate and the most common source of misclassification, because so many patients are already on diuretics when the effusion is tapped.
When diuretics confuse the result
Diuretics pull water out of the pleural space faster than protein and LDH, so the fluid becomes more concentrated and can cross Light's thresholds. Two follow-up checks help:
- Serum–effusion albumin gradient. Serum albumin minus fluid albumin. Roth and colleagues (Chest, 1990) used above 1.2 g/dL for a transudate; it correctly classified 57 of 59 effusions, and Light's criteria had misclassified five heart failure effusions, four after diuretics.
- Serum–fluid protein gradient. Serum protein minus fluid protein. A gradient above 3.1 g/dL favours a transudate, and the table from Light's 2002 review recommends it when Light's criteria are met but no value is more than 15% above its cutoff.
The calculator shows both gradients and the 15% borderline flag so you can see when a second look is worth it.
Common mistakes
- Using a fixed 200 U/L LDH cutoff. Use two-thirds of your own lab's upper limit of normal.
- Mismatched units. Protein ratios cancel units, but the gradients need g/dL; the calculator converts from g/L.
- Serum taken days apart. Ratios need serum drawn close to the thoracentesis.
- Stopping at "exudate". The criteria classify fluid; they do not name a cause.
- Ignoring diuretics. A borderline exudate in a patient on diuretics is often a transudate.
More clinical tools: A/G ratio, anion gap, Winter's formula and the QTc calculator.
For healthcare professionals and students only. Light's criteria classify pleural fluid; they do not diagnose its cause, and this tool offers no medical advice. Breathlessness, chest pain or fever with an effusion needs prompt clinical review. Patients should see their own clinician.
Frequently asked questions
What are Light's criteria?
Light's criteria classify a pleural effusion as an exudate if at least one of three is true: pleural fluid protein ÷ serum protein is greater than 0.5; pleural fluid LDH ÷ serum LDH is greater than 0.6; or pleural fluid LDH is greater than two-thirds of the upper limit of normal for serum LDH. If none is met, the effusion is a transudate.
What is the difference between a transudate and an exudate?
A transudate forms when pressure or protein balance changes push fluid across normal pleura, as in heart failure or cirrhosis. An exudate forms when the pleura itself is inflamed, infected or invaded, as in pneumonia, cancer or tuberculosis. The distinction decides how far the search for a cause needs to go.
How accurate are Light's criteria?
Very sensitive, less specific. In the table adapted from Light's 2002 NEJM review, the combined criteria had 98% sensitivity and 77% specificity for an exudate. They almost never miss an exudate, but they label some transudates as exudates.
Why do Light's criteria misclassify heart failure effusions?
Diuretics remove water from the pleural space faster than protein and LDH, which concentrates them. Roth and colleagues (1990) found Light's criteria misclassified five heart failure effusions as exudates, four of them in patients who had had diuretics.
What is the serum–effusion albumin gradient?
Serum albumin minus pleural fluid albumin. In Roth and colleagues' 1990 study, a gradient above 1.2 g/dL indicated a transudate and 1.2 g/dL or less an exudate; it classified 57 of 59 patients correctly and was more specific than Light's criteria in chronic heart failure.
When should I use the protein gradient instead?
The table from Light's 2002 review lists a serum minus pleural fluid protein gradient of more than 3.1 g/dL as favouring a transudate. It is preferred in patients started on diuretics after the effusion developed who meet Light's criteria, but where no measurement is more than 15% above its cutoff.
Did Light's original paper use 200 U/L for LDH?
Yes. The 1972 paper used a pleural fluid LDH above 200 IU as the third criterion, which was two-thirds of that laboratory's upper limit of normal. Later work, including Light's 2002 review, expresses it as two-thirds of the upper limit of normal so it works across different LDH assays.
What units do I need?
Protein and albumin can be in g/dL or g/L because the ratios cancel the units; the calculator converts to g/dL for the gradients. LDH must be in the same unit (U/L) for fluid, serum and the upper limit of normal.
Should serum and pleural samples be taken at the same time?
Yes. The ratios assume serum and fluid reflect the same moment, so draw blood close to the time of thoracentesis.
Does an exudate tell you the cause?
No. It tells you the pleura is involved. Further tests such as cell count, glucose, pH, microbiology and cytology are used to find the cause.
Sources & method
- Light RW, MacGregor MI, Luchsinger PC, Ball WC Jr (1972), Annals of Internal Medicine 77:507–513 — Pleural effusions: the diagnostic separation of transudates and exudates
- Light RW (2002), New England Journal of Medicine 346:1971–1977 — Clinical practice: pleural effusion
- MSD Manual Professional Edition — Criteria for identifying exudative pleural effusions (table adapted from Light, NEJM 2002)
- Roth BJ, O'Meara TF, Cragun WH (1990), Chest 98:546–549 — The serum-effusion albumin gradient in the evaluation of pleural effusions
- Roberts ME, Rahman NM, Maskell NA et al. (2023), Thorax — British Thoracic Society guideline for pleural disease
Results are estimates for general information. Found an error? It helps everyone — see our methodology.