Serum and plasma are both derived from blood, but they differ in one critical respect: serum is the liquid left after blood has been allowed to clot and the clot is removed, while plasma is obtained by preventing clotting with an anticoagulant and then spinning off the cells. That single distinction, whether the clotting process is allowed to happen, ripples through nearly every corner of diagnostic medicine. It affects which analytes you can accurately measure, how quickly a lab can report results, and whether a sample is suitable for molecular testing. The choice between the two is rarely arbitrary, and getting it wrong can produce misleading results.
How Each Sample Type Is Made
Blood drawn into a plain tube or a tube containing a clot activator (such as silica particles or thrombin) is allowed to sit until the coagulation cascade runs its course. Fibrinogen converts to fibrin, platelets aggregate into a mesh, and the whole mass contracts into a solid clot. The tube is then centrifuged, and the clear yellowish fluid that separates from the clot is serum. Because the clotting factors have been consumed in forming the clot, serum lacks fibrinogen and several other coagulation proteins.
Plasma collection takes the opposite approach. Blood is drawn into a tube containing an anticoagulant that halts coagulation before it begins. Different anticoagulants work through different mechanisms. Sodium citrate and oxalate bind free calcium, which the clotting cascade needs to proceed. EDTA chelates calcium as well. Heparin works differently: it amplifies the activity of antithrombin, a natural inhibitor that shuts down thrombin and other clotting enzymes.1PubMed. Clot activators and anticoagulant additives for blood collection. A critical review on behalf of COLABIOCLI WG-PRE-LATAM After centrifugation, the liquid above the packed cells is plasma, which still contains fibrinogen and the full complement of clotting factors.
The anticoagulant choice is not interchangeable. Each one is matched to the test being performed, and using the wrong tube can ruin the sample or skew the numbers. Laboratories color-code their collection tubes for exactly this reason: lavender caps for EDTA, light blue for citrate, green for heparin, red or gold for serum.
Why Potassium Readings Differ Between Serum and Plasma
One of the best-known practical differences involves potassium. When blood clots in a serum tube, platelets and red blood cells release potassium into the surrounding fluid. In a healthy person, this in vitro release is small enough that serum and plasma potassium values stay close. But in patients with very high platelet counts or white blood cell counts, the release during clotting can push the serum potassium reading well above what the patient’s actual circulating level is. This artifact is called pseudohyperkalemia, and it is defined as a serum potassium concentration exceeding the paired plasma potassium by more than 0.4 mmol/L when both samples are drawn at the same time, kept at room temperature, and tested within an hour.2PubMed Central. Pseudohyperkalemia in serum: a new insight into an old phenomenon3PubMed Central. Pseudohyperkalemia in Serum and Plasma: The Phenomena and Its Clinical Implications
The clinical stakes are real. A falsely elevated potassium level can trigger unnecessary treatment, including urgent interventions to lower potassium that the patient never actually needed. Many hospitals have shifted their routine chemistry panels to plasma (usually lithium-heparin) tubes partly to avoid this issue. When a clinician sees an unexpectedly high potassium result on a serum sample, especially in a patient with leukemia or a myeloproliferative disorder, the standard next step is to redraw into a plasma tube and compare.
Coagulation Testing Requires Plasma
Serum is categorically unusable for coagulation studies like the prothrombin time (PT) or the activated partial thromboplastin time (aPTT). These tests work by re-starting the clotting cascade under controlled conditions and measuring how long it takes. If the clotting factors have already been consumed in forming a clot, as they are in serum, there is nothing left to measure.
The standard collection tube for coagulation work contains 3.2% sodium citrate at a ratio of one part anticoagulant to nine parts whole blood.4PubMed Central. Evaluation of Citrated Plasma After Thawing for Routine Coagulation Testing Citrate binds calcium, putting the coagulation cascade in a holding pattern. When the lab is ready to run the test, calcium chloride is added back to the plasma, restoring the conditions needed for clotting to proceed.5Laboratory Medicine. Coagulation Testing in the Core Laboratory Underfilling or overfilling the tube changes the citrate-to-blood ratio, which shifts the results. That is why coagulation tubes carry strict fill-line requirements that laboratories enforce rigorously.6Mayo Clinic Proceedings. Approach to the Diagnosis and Management of Bleeding Disorders
EDTA Contamination and Cross-Tube Errors
EDTA is the anticoagulant of choice for complete blood counts and many hematology tests. It is excellent at preserving cell morphology. But when even a small amount of EDTA leaks into a chemistry sample, whether from a poorly ordered draw sequence, a tube that splashes, or an accidental tube swap, it wreaks havoc on several analytes. Calcium plummets because EDTA chelates it. Potassium rises because EDTA salts contain potassium. Alkaline phosphatase drops because it is a zinc- and magnesium-dependent enzyme and EDTA strips those metal cofactors away.7PubMed. Effects of contamination of blood specimens with liquid potassium-EDTA anticoagulant Even transaminases, lactate dehydrogenase, creatine kinase, and amylase can be affected.
The magnitude of the interference depends on how much EDTA gets in. One study found that serum potassium rose and calcium fell beyond clinically meaningful thresholds at EDTA concentrations as low as 0.17 mmol/L, while alkaline phosphatase was more resilient, not crossing its threshold until much higher concentrations.8PubMed. Managing ethylenediaminetetraacetic acid (EDTA) interference in EDTA contaminated samples – selectivity in reporting analytes The vulnerability of a given test also depends on the assay method the lab uses. Magnesium measured by one enzymatic method resists EDTA well, while magnesium measured by a different dye-binding method is knocked off course at lower EDTA levels.9PubMed. Assay-Dependent Effects of EDTA Contamination on Magnesium and Iron Iron behaves similarly: one detection chemistry tolerates EDTA fine, another does not. This assay-method dependence is one reason two hospitals can have different policies about which contaminated results are salvageable.
Cell-Free DNA and Molecular Testing
In molecular diagnostics, the serum-versus-plasma decision matters a great deal, and the “better” choice depends on what you are looking for. Plasma is strongly preferred for cell-free DNA (cfDNA) testing, including non-invasive prenatal screening and liquid biopsies for cancer. The reason traces back to clotting: when blood coagulates in a serum tube, white blood cells trapped in the clot break open and spill their genomic DNA into the surrounding fluid. That burst of normal DNA dilutes the signal from the fragments of fetal or tumor DNA the test is trying to detect.
Studies comparing matched serum and plasma samples consistently show higher total cfDNA concentrations in serum, but the extra DNA is overwhelmingly from white blood cells and is not diagnostically useful.10Clinica Chimica Acta. Circulating cell free DNA: Preanalytical considerations Research on prenatal screening found that the fetal fraction of cfDNA was significantly lower in serum than in plasma, and the size distribution of cfDNA fragments in serum shifted toward longer fragments, consistent with contamination by intact genomic DNA released from blood cells.11PubMed. Cell-free DNA in maternal plasma and serum: A comparison of quantity, quality and tissue origin using genomic and epigenomic approaches Methylation-based tissue mapping confirmed that the excess DNA in serum came predominantly from neutrophils and B cells. Different serum tube formulations also affected the degree of contamination, meaning that switching brands of tubes could shift results.
The anticoagulant in the plasma tube matters too. Heparin, while useful for chemistry panels, is a known inhibitor of PCR, the amplification reaction at the heart of most molecular tests.12PubMed Central. An optimized sensitive method for quantitation of DNA/RNA viruses in heparinized and cryopreserved plasma13PubMed Central. A simple method to overcome the inhibitory effect of heparin on DNA amplification EDTA plasma is generally the preferred matrix for PCR-based molecular testing because EDTA does not interfere with the polymerase enzyme the way heparin does. Specialized cfDNA collection tubes that contain EDTA plus additional cell-stabilizing agents have become the standard for liquid biopsy workflows, holding cells intact during transport so their DNA does not leak into the plasma.
Antibody and Serology Testing
For many antibody-based tests, serum and plasma are functionally interchangeable. A study comparing antibody titers to mycobacterial antigens in matched serum and plasma samples from both HIV-positive and HIV-negative patients found correlations above 0.89 across all antigens tested, concluding that either sample type could be used in the same assay without meaningful differences.14PubMed Central. Correlation between serum and plasma antibody titers to mycobacterial antigens More recently, a comparison of SARS-CoV-2 serology using 124 paired serum and plasma samples found no meaningful difference in IgG, IgM, or IgA antibody concentrations or in functional neutralization results across multiple assay platforms.15PubMed. Comparative performance of serum and plasma samples in SARS-CoV-2 serology and neutralization assays
Serology has historically defaulted to serum, in part because early antibody assays were validated that way and because serum avoids any risk of anticoagulant interference. But the practical interchangeability confirmed across infections ranging from tuberculosis to COVID-19 gives labs flexibility. In large population studies or biobank-based research, the ability to use plasma instead of serum simplifies logistics and can reduce costs.
Platelet-Derived Biomarkers and Cytokines
Not all analytes are as forgiving. Biomarkers stored inside platelets present a predictable challenge: anything that platelets release when they activate during clotting will be higher in serum than in plasma. The textbook example is vascular endothelial growth factor (VEGF), a protein involved in blood vessel formation that is heavily concentrated in platelets. Research showed that the large difference between VEGF levels in serum and plasma is directly due to platelets releasing their VEGF stores during coagulation.16PubMed Central. Release of the angiogenic cytokine vascular endothelial growth factor (VEGF) from platelets: significance for VEGF measurements and cancer biology
This matters for cancer research, where VEGF levels are sometimes used to track tumor angiogenesis. If you measure VEGF in serum, you are mostly measuring what platelets dumped during clotting, not the circulating level that reflects what the tumor is doing. Plasma gives a more accurate picture of in vivo VEGF, but the sample must be processed carefully to avoid even partial platelet activation during handling. The same principle applies to other platelet-derived growth factors and cytokines. When designing biomarker studies, researchers systematically evaluate which matrix gives the most biologically informative measurement for each analyte, a process that sometimes yields surprises depending on how much of the biomarker lives inside cells versus circulates freely.
Turnaround Time in the Lab
Plasma has a practical speed advantage. Serum tubes need time for the blood to clot, typically 30 minutes at room temperature, before centrifugation can happen. Plasma tubes skip this wait entirely: the anticoagulant prevents clotting, so the tube can go straight into the centrifuge. In emergency departments and critical care settings, those 30 minutes matter.
A study at a large university hospital that switched routine chemistry testing from serum to plasma tubes with mechanical separators found that the average time from sample receipt to potassium result dropped from about 36 minutes to about 30 minutes.17The Journal of Applied Laboratory Medicine. Improved Sample Quality and Decreased Turnaround Time When Using Plasma Blood Collection Tubes with a Mechanical Separator in a Large University Hospital A six-minute improvement sounds modest in isolation, but across thousands of daily samples it reshapes workflow, and for critically ill patients a faster potassium or troponin result can influence treatment timing.
Centrifugation itself is a bottleneck that has been well studied. Standard protocols call for about 10 minutes at roughly 1,200 g on a swing-out rotor, with some guidelines suggesting higher forces or longer spins to better clear platelets from the supernatant.18Laboratory Medicine. Preparation of a Quality Sample: Effect of Centrifugation Time on Stat Clinical Chemistry Testing Most routine chemistry analytes tolerate reduced centrifugation times at higher g-forces without meaningful changes. The notable exception in one evaluation was lactate dehydrogenase, which showed a roughly 6% positive bias when centrifugation was shortened, likely because incomplete platelet removal left cellular enzymes in the supernatant.19PubMed. Evaluation of a reduced centrifugation time and higher centrifugal force on various general chemistry and immunochemistry analytes in plasma and serum
Freeze-Thaw Stability and Biobanking
Biobanks that store thousands of samples for future research need to know how well analytes survive repeated freezing and thawing. Some degradation is inevitable, but most routine chemistry analytes hold up reasonably well through several cycles. A study of 42 analytes in serum and plasma found that the vast majority remained stable through four freeze-thaw cycles at minus 80 degrees Celsius, with HDL cholesterol, chloride, and sodium being among the exceptions that drifted beyond desirable limits.20PubMed. Sample stability of forty-two analytes in plasma or serum pools after one to four repeated -80 °C freeze-thaw cycles Another study tracking stored serum over three months found that albumin, total protein, and blood urea nitrogen showed significant changes with time, while liver enzymes, glucose, creatinine, cholesterol, and triglycerides remained stable even after ten freeze-thaw cycles.21Biochemia Medica. The effect of storage time and freeze-thaw cycles on the stability of serum samples
Cytokines and protein biomarkers are more temperamental. Research examining eight biomarkers in matched plasma and serum samples found that some, like IL-8 and VEGF-R2, stayed put through five freeze-thaw cycles in both matrices. Others behaved differently depending on the matrix. IL-15 and IL-17A were stable in plasma but shifted in serum. MMP-7 and VEGF tended to rise with repeated thawing in both, though the increase was more limited in serum.22Osong Public Health and Research Perspectives. Effect of Repeated Freezing and Thawing on Biomarker Stability in Plasma and Serum Samples These matrix-specific stability profiles mean biobank designers cannot simply assume that a finding validated in serum will hold when the stored samples happen to be plasma, or vice versa.
Metabolomics Differences
Metabolomics, the large-scale study of small molecules in biological samples, is particularly sensitive to the serum-versus-plasma distinction. A head-to-head comparison using both NMR spectroscopy and mass spectrometry found that after accounting for normal person-to-person variation, several metabolite classes reliably distinguished serum from plasma. Lipoproteins, lipids carried in VLDL and LDL particles, lactate, glutamine, and glucose all differed between the two matrices.23PubMed Central. A comparison of human serum and plasma metabolites using untargeted (1)H NMR spectroscopy and UPLC-MS Some of these shifts reflect the metabolic activity of cells during clotting: red blood cells and platelets consume glucose and produce lactate, so serum tends to have lower glucose and higher lactate than plasma from the same draw. Others reflect the release or binding of molecules during clot formation.
The practical consequence is that metabolomics studies must commit to one matrix and stick with it across all samples in a study. Mixing serum and plasma introduces systematic variation that can swamp the biological differences the study is trying to detect. Published metabolomics datasets that do not specify the matrix, or that pooled both types, are viewed skeptically by the field.
Therapeutic Drug Monitoring and Gel Separator Tubes
Both serum and plasma are used for measuring drug levels, but the tubes they come in introduce a subtle problem. Many modern collection tubes contain a gel separator that forms a physical barrier between the liquid and the cell layer during centrifugation. The gel is convenient: it stabilizes the sample and makes it easier to pipette. But gel separators can absorb certain drugs out of the sample over time, lowering the measured concentration.
This absorption is not uniform. Lipophilic (fat-soluble) drugs are more prone to being soaked up by the gel because the polymer is itself lipophilic. Research tracking a range of opioids found that synthetic opioids, which tend to be more lipophilic than their naturally derived counterparts, showed the most significant drop in measured concentration after storage in gel-separator tubes.24Journal of Analytical Toxicology. Adsorption of Therapeutic and Recreational Drugs During Prolonged Storage of Plasma Samples in Gel Separator Tubes Early work on antiepileptic drugs demonstrated the same phenomenon: phenytoin and phenobarbital concentrations fell in a time- and volume-dependent manner in gel tubes, and the drugs could be recovered by extracting the gel with solvent, confirming that absorption was the mechanism.25PubMed. Absorption of therapeutic drugs by barrier gels in serum separator blood collection tubes. Volume- and time-dependent reduction in total and free drug concentrations
Among antihypertensive medications, a study of 21 drugs found that most were stable in gel tubes for at least 72 hours. But verapamil dropped by more than 40%, lercanidipine by about 30%, and diltiazem by roughly 19% compared to samples in non-gel tubes over the same period.26PubMed Central. Stability of 21 Antihypertensive Drugs in Serum Collected in Standard (Nongel) Serum Tubes Versus Tubes Containing a Gel Separator For drugs with narrow therapeutic windows, where a small change in measured concentration could mean the difference between “in range” and “subtherapeutic,” this absorption matters. Labs that perform therapeutic drug monitoring often specify non-gel tubes or require prompt separation and aliquoting to minimize contact time between the gel and the sample.
Veterinary Applications
The serum-versus-plasma question extends to animal medicine, where the same fundamental chemistry applies but some species-specific wrinkles emerge. In cattle, for instance, some copper in the blood is known to get trapped in the clot during coagulation, meaning serum copper values may underestimate what is actually circulating.27PubMed Central. Determination of Essential and Toxic Elements in Cattle Blood: Serum vs Plasma For trace mineral assessments in livestock, where the clinical question is whether animals are deficient, that systematic undercount in serum can tip a borderline result in the wrong direction. In poultry, however, a study comparing multiple tube types and centrifugation speeds found no meaningful differences in routine biochemistry parameters like glucose, cholesterol, total protein, or calcium between serum and the various plasma preparations tested.28Research and Practice in Veterinary and Animal Science. Effect of Centrifugation Speed and Blood Tubes with Different Contents on Certain Serum/Plasma Biochemical Parameters in Cocks The takeaway for veterinary practice is the same as for human medicine: the choice depends on the analyte and the clinical question, not on a blanket preference for one matrix over the other.