Albumin Blood Test: Low, High & Normal Range Explained

Albumin is made by the liver and is the most abundant protein in plasma. It helps retain fluid within the circulation and carries substances including bilirubin, calcium, hormones, fatty acids, vitamins and many medicines. The laboratory test measures its concentration in serum or plasma, usually by a dye-binding method.^{See reference 1,See reference 2,See reference 6}

A concentration depends on both the amount of albumin and the volume and location of fluid around it. A value can therefore fall because albumin production falls, albumin is lost or redistributed, breakdown increases, or the blood is diluted. It can rise when plasma water falls. One number cannot identify which process occurred.^{See reference 2,See reference 6,See reference 7,See reference 8}

Albumin is useful context within a liver panel or metabolic panel, but it is not a stand-alone liver-function score, kidney diagnosis, nutrition assessment, hydration prescription or prognosis. A normal result can coexist with important disease, including early acute liver injury or kidney disease with urine albumin loss that has not yet lowered blood albumin.^{See reference 2,See reference 4,See reference 10,See reference 11}

Albumin is commonly reported in grams per litre (g/L) or grams per decilitre (g/dL). The conversion is exact: multiply g/dL by 10 to get g/L, or divide g/L by 10 to get g/dL. For example, 3.7 g/dL is 37 g/L. Keep the laboratory range in the same unit as the result before comparing them.^{See reference 2,See reference 8,See reference 9,See reference 13}

The test uses venous blood and usually needs no special preparation on its own. A combined panel may require fasting for another analyte. Tell the clinician or laboratory about medicines and recent infusions, but do not stop medicines because of this page.^{See reference 1,See reference 8,See reference 9,See reference 13}

Serum and plasma methods can both be used. Record the specimen, laboratory and interval printed on the report; for meaningful trends, use the same laboratory and method when practical rather than converting one assay result into another.^{See reference 2,See reference 5,See reference 8,See reference 9}

Albumin is one marker of liver synthetic function, while ALT, AST, ALP and GGT mainly reflect patterns of injury or cholestasis. BSG guidance recommends interpreting abnormal liver blood tests with previous results, history and current condition; the size of an abnormality alone does not determine its clinical importance.^{See reference 4}

Because albumin turns over relatively slowly, it may remain normal early in even severe acute liver disease. In known chronic liver disease, a falling value can be important, but inflammation, kidney or gut loss and dilution can produce the same direction of change.^{See reference 2,See reference 4,See reference 6,See reference 7}

A normal albumin result therefore does not rule out steatotic liver disease, hepatitis, fibrosis or an acute liver problem, and a low result does not prove that the liver is the cause. Bilirubin, INR, platelets, enzymes, imaging and the clinical story answer different parts of the question.^{See reference 1,See reference 2,See reference 4}

Healthy kidney filters retain most albumin in the blood. Glomerular damage can allow protein to enter urine. When loss is large enough, blood albumin may fall and swelling, high lipids and other features can form the nephrotic-syndrome pattern.^{See reference 5,See reference 10,See reference 11}

Urine albumin or UACR can be abnormal while serum albumin remains normal, especially earlier in kidney disease. Conversely, low serum albumin can come from inflammation, liver dysfunction, dilution or gut/skin loss without urine albumin being the cause.^{See reference 2,See reference 10,See reference 11}

Do not compare their numbers directly: serum albumin is commonly g/L or g/dL, while UACR is a urine ratio such as mg/g or mg/mmol. A kidney interpretation needs urine protein or UACR plus urinalysis, creatinine/eGFR, blood pressure, symptoms and trend.^{See reference 5,See reference 10,See reference 11}

ASPEN's position paper states that serum albumin and prealbumin should not be used as proxy measures of total body protein or muscle mass and should not be used alone as nutrition markers. Their concentrations commonly fall with inflammation regardless of underlying nutrition status.^{See reference 3}

That does not mean nutrition never matters. Poor intake, malabsorption and disease-related malnutrition can coexist with inflammation and may contribute, but assessment uses weight and intake history, physical findings, muscle and fat loss, function, symptoms and disease context—not albumin alone.^{See reference 1,See reference 2,See reference 3,See reference 7}

Protein powder cannot correct urinary loss, advanced liver dysfunction, capillary leakage or fluid overload. Increasing protein without understanding kidney, liver, heart or fluid status can also be inappropriate. Use a clinician or dietitian for an individualized nutrition assessment when intake or weight loss is a concern.^{See reference 3,See reference 5,See reference 7,See reference 10}

A genuinely high serum albumin concentration is uncommon. Professional laboratory guidance points mainly to water depletion and recent plasma or albumin-containing infusion. Vomiting, diarrhoea, heavy sweating or other fluid loss can concentrate the blood, but the cause and safe fluid plan depend on the person and illness.^{See reference 1,See reference 2,See reference 8,See reference 14}

Prolonged tourniquet time and upright collection can also make a result look higher. Compare the result with sodium, urea/BUN, creatinine, total protein, symptoms and collection conditions, and repeat under stable conditions when clinically appropriate.^{See reference 2,See reference 8,See reference 13}

A high result is not a score of dietary protein and does not justify forced water intake. People with heart, kidney or liver conditions may have fluid restrictions or complex fluid needs; severe dehydration or shock symptoms require urgent assessment rather than a web hydration target.^{See reference 2,See reference 8,See reference 14}

Some calcium circulates bound to albumin, so laboratories have historically used equations to report albumin-adjusted or “corrected” calcium. These formulas do not measure biologically active ionized calcium and their performance changes with illness, pH, kidney function, albumin assay and patient population.^{See reference 2,See reference 12}

In 2026, a joint position statement from the International Osteoporosis Foundation working group, IFCC Committee on Bone Metabolism and EFLM Committee on CKD recommended that laboratories stop routine reporting of albumin-adjusted calcium. It recommends total calcium as the default and ionized calcium when decisions depend on calcium status or total calcium is likely to be unreliable.^{See reference 12}

Do not calculate a personal corrected-calcium result from this page or use albumin to overrule a calcium plan. Keep the measured total and ionized calcium values, albumin method and clinical setting together, and ask which measure should guide the decision.^{See reference 12}

There is no one retest interval for every abnormal albumin result. Acute symptoms, a large change, known liver or kidney disease, pregnancy, recent surgery, burns, infusion or fluid loss can make the correct timing much faster than a stable, mild, unexpected result.^{See reference 1,See reference 2,See reference 4,See reference 7,See reference 13}

For a trend, compare the original units, laboratory interval, specimen and method. Note posture, difficult/prolonged collection, fasting required for companion tests, medicines, pregnancy, illness, weight and recent IV fluids. ACB reports a 12% critical difference in its monograph, but that is method/population context—not a universal personal change threshold.^{See reference 2,See reference 8,See reference 9}

Repeating albumin alone may confirm the concentration but cannot locate the cause. The useful companion tests depend on the suspected mechanism: liver panel and INR, urine protein/UACR and kidney function, CRP, total protein/globulin, blood count, digestive evaluation, or a formal nutrition assessment.^{See reference 3,See reference 4,See reference 5,See reference 10,See reference 11}