Protein testing spans a surprisingly wide range of techniques, from a simple paper dipstick dipped in urine to mass spectrometers that can identify thousands of individual proteins in a single sample. The method that makes sense depends entirely on the question being asked: a food scientist checking whether a batch of soy flour meets its label claim uses a completely different approach than a nephrologist tracking kidney disease or a biochemist purifying an enzyme. What unites all these methods is that they exploit some physical or chemical property of proteins — their nitrogen content, their ability to bind dyes, their absorption of light, or their unique molecular shape — to either measure how much protein is present or identify which protein it is.
Dye-Binding and Color-Change Assays
The most common way to measure total protein in a research lab is to mix a sample with a chemical reagent and watch the color change. These colorimetric assays are cheap, fast, and require nothing more sophisticated than a device that reads how much light passes through a colored solution. Several variations exist, each with its own strengths and blind spots.
The Bradford assay is probably the most widely used. It works by adding a dye called Coomassie Brilliant Blue G-250 to a protein solution. The dye binds to proteins and shifts from a brownish-red color to blue, and the intensity of the blue color reflects how much protein is present. The binding happens mainly through interactions with the amino acid arginine, along with weaker contributions from a handful of other amino acids. Van der Waals forces and hydrophobic interactions drive the reaction.1PubMed. Mechanism of dye response and interference in the Bradford protein assay The Bradford assay is quick and sensitive, but because it depends on specific amino acids, it responds differently to different proteins. A protein rich in arginine will produce a stronger signal per milligram than one that is not, which can skew results if your sample does not resemble the protein you used to build your standard curve. Detergents are the most problematic source of interference, and certain small peptides and transfection reagents can also distort readings.2ChemRxiv. Shifting focus in the Bradford assay: interfering compounds re-examined
The BCA (bicinchoninic acid) assay takes a different chemical route. Proteins reduce copper ions in an alkaline solution, and the resulting copper then reacts with BCA to produce a purple color. The BCA assay tolerates most detergents at moderate concentrations, which makes it popular when samples contain surfactants. Its Achilles’ heel is reducing agents — chemicals like DTT and beta-mercaptoethanol that are commonly used to break disulfide bonds in proteins. These reducing agents react directly with the copper reagent and inflate the apparent protein concentration.3The FASEB Journal. Development of a Detergent and Reducing Agent Compatible BCA Protein Assay
The biuret assay is the simplest and oldest of the group. It relies on copper ions forming chelates with peptide bonds under alkaline conditions. Because it responds to peptide bonds rather than specific amino acids, it is less biased by protein composition than the Bradford assay. The biuret method is frequently cited as a reference method for total protein measurement, particularly in veterinary and clinical biochemistry.4PubMed Central. Comparison of biuret and refractometery method for serum total protein measurements in cattle and goat Its main drawback is lower sensitivity — you need a fairly concentrated sample to get a reliable reading.
The Lowry assay combines the biuret reaction with a second step involving the Folin-Ciocalteu reagent, which reacts with aromatic amino acids like tryptophan and tyrosine to produce a deep blue color. This two-step process makes the Lowry more sensitive than the biuret alone, but also more time-consuming and more prone to interference from common lab chemicals. The underlying chemistry of that second step is still not fully understood, even after decades of use.
Across all colorimetric assays, accuracy depends on how closely your sample’s amino acid makeup resembles the standard protein used to build the calibration curve — usually bovine serum albumin.5PubMed Central. Evaluation of colorimetric assays for analyzing reductively methylated proteins: Biases and mechanistic insights If your protein of interest is very different from albumin in composition, the number you get back could be off by a meaningful amount. This is a consistent limitation of the whole family of tests, not a flaw unique to any one assay.
Measuring Protein with UV Light
If you need a fast estimate and have a reasonably pure protein solution, you can skip reagents entirely and just shine ultraviolet light through your sample. Proteins absorb UV light at 280 nanometers because of their aromatic amino acids, primarily tryptophan and tyrosine. The amount of light absorbed is proportional to the protein concentration, and if you know the protein’s sequence, you can calculate its expected absorption strength in advance. This makes UV absorbance a convenient, non-destructive method — the sample is unchanged afterward and can still be used for experiments.6PubMed Central. Sequence-specific determination of protein and peptide concentrations by absorbance at 205 nm
The catch is that the method only works for proteins containing tryptophan or tyrosine. A protein that lacks both of these amino acids will be essentially invisible at 280 nm. Nucleic acids also absorb strongly in this UV range, so a sample contaminated with DNA or RNA will give an inflated protein reading. For crude cell extracts or unpurified samples, UV absorbance is unreliable. But for checking the concentration of a purified protein in a clean buffer, it is hard to beat for speed and convenience.
How Food Protein Content Is Measured
The protein content printed on a food label is almost never measured by directly detecting protein. Instead, it is measured indirectly through nitrogen. The Kjeldahl method, developed in the 1880s, remains the internationally accepted reference method for protein in food and feed. The sample is digested in hot concentrated sulfuric acid with a catalyst, converting bound nitrogen into ammonium sulfate. The ammonia is then released, distilled, and measured by titration. The resulting nitrogen value is multiplied by a conversion factor to estimate protein content.7PubMed Central. Application of the Ninhydrin Reaction for Quantification of Total Protein Contents: Establishment of Conversion Formulas
That conversion factor is where things get interesting. The traditional default is 6.25, based on an old assumption that all proteins contain 16% nitrogen. Amino acid analyses have since shown this is too high for most foods — different proteins have different nitrogen contents depending on their amino acid profiles, and foods also contain non-protein nitrogen from sources like nucleic acids, free amino acids, and other compounds. Specific conversion factors have been proposed for many food categories. Soy, for instance, has a factor of 5.71 derived from the nitrogen content of its main storage protein, but the soy industry and many regulatory bodies still use 6.25, partly because switching would change the apparent protein content on labels and affect product valuation.8Journal of the American Oil Chemists’ Society. Calculation of Nitrogen‐to‐Protein Conversion Factors: A Review with a Focus on Soy Protein
The Dumas combustion method is a faster alternative to Kjeldahl. Instead of wet chemistry, it burns the sample at high temperature and measures the nitrogen in the resulting gases. Results between the two methods are highly correlated across cereals, meats, seafood, and vegetables.9Analytical Science and Technology. A comparison study of crude protein contents obtained utilizing the Kjeldahl method and Dumas combustion method in foods The Dumas method avoids the hazardous chemicals required by Kjeldahl, produces results in minutes rather than hours, and is increasingly replacing Kjeldahl in routine labs. Both methods, however, share the same fundamental weakness: they measure nitrogen, not protein. Anything in the sample that contains nitrogen but is not protein will inflate the result.
The Melamine Scandal and Nitrogen-Based Fraud
The fact that Kjeldahl and Dumas methods cannot distinguish protein nitrogen from other nitrogen creates a vulnerability to deliberate fraud. The most notorious example was the 2008 melamine contamination of Chinese milk products. Melamine is a nitrogen-rich industrial chemical that, when added to diluted milk, raises the apparent protein content without adding any actual protein. The Kjeldahl test simply reads the extra nitrogen and reports a higher number.
Closing this loophole requires a separate step: measuring the non-protein nitrogen (NPN) fraction. One approach is to precipitate the actual proteins out of solution using an acid like trichloroacetic acid or tannic acid, then run the Kjeldahl test on the remaining liquid. If the non-protein nitrogen is unusually high, that signals possible adulteration. Research on milk powders has shown that tannic acid precipitation and molecular-weight filtration are both practical screening methods for detecting a variety of nitrogen-rich adulterants in routine food analysis labs.10Elsevier (International Dairy Journal). Non-protein nitrogen determination: A screening tool for nitrogenous compound adulteration of milk powder
Urine Protein Tests in Clinical Medicine
When doctors test for protein, the context is often urine. Healthy kidneys filter waste but retain protein; finding protein in urine — proteinuria — can signal kidney damage, preeclampsia during pregnancy, or systemic diseases. The simplest screening tool is the urine dipstick, a paper strip impregnated with a pH-sensitive dye. When proteins in the urine bind to the dye’s anionic form, the color shifts from yellow toward green or blue, even though the pH has not changed. These dipsticks are strongly biased toward albumin because albumin carries more positively charged amino groups than other proteins.11Elsevier / American Journal of Kidney Diseases. Review Methods for Diagnosing Proteinuria—When to Use Which Test and Why: A Review That bias means a dipstick can miss elevated levels of other proteins, such as immunoglobulin light chains, which matter in conditions like multiple myeloma.
For a more quantitative picture, doctors can order a 24-hour urine collection, which measures the total protein excreted over a full day. This is considered the gold standard for quantifying proteinuria, but it is cumbersome — asking a patient to collect every drop of urine for 24 hours is inconvenient and prone to errors from missed or spilled samples. The protein-to-creatinine ratio (PCR), measured on a single random urine sample, has become a widely used alternative. It adjusts for how concentrated the urine is by dividing the protein level by the creatinine level, giving a snapshot that correlates reasonably well with 24-hour collection. One prospective study found that the correlation between spot PCR and 24-hour protein was strong, with sensitivity above 90% for detecting significant proteinuria.12PubMed Central. The urine protein/creatinine ratio as a reliable indicator of 24-h urine protein excretion across different levels of renal function and proteinuria: the TUNARI prospective study
The agreement between spot PCR and 24-hour collection weakens at higher levels of protein loss. When proteinuria is heavy, the spot test tends to become unreliable, and clinicians treating diseases like lupus nephritis or nephrotic syndrome still rely on full 24-hour collections for accurate monitoring.13PubMed. Can spot urine protein/creatinine ratio replace 24 h urine protein in usual clinical nephrology? So if you have been asked to do a 24-hour collection despite the existence of a simpler spot test, there is usually a clinical reason for it.14Journal of General Practice and Emergency Medicine of Nepal. Correlation between spot urine protein creatinine ratio and 24-hour urinary protein in systemic lupus erythematosus patients with proteinuria
Blood Protein Tests and What the Numbers Mean
A routine blood panel often includes total protein, albumin, and globulin levels. Total serum protein reflects the combined concentration of all proteins circulating in your blood, with albumin and globulins making up the vast majority. Albumin, produced by the liver, keeps fluid from leaking out of blood vessels and carries hormones, drugs, and nutrients. Globulins include immune antibodies, transport proteins, and clotting factors. The ratio of albumin to globulin (the A/G ratio) can flag problems: a low ratio may suggest chronic inflammation, liver disease, or overproduction of immunoglobulins. In COVID-19 pneumonia patients, for instance, falling albumin levels and a declining A/G ratio were independently associated with worse clinical outcomes, including longer hospital stays.15PubMed Central. Serum Albumin, Globulin and Albumin–Globulin Ratios as Biomarkers of Clinical Outcomes in COVID-19 Pneumonia A low A/G ratio has also been linked to increased risk of interstitial lung disease in patients with Sjögren’s syndrome.16PubMed. Low albumin-to-globulin ratio is associated with interstitial lung disease in primary Sjögren’s syndrome
When a doctor suspects that one particular protein is being overproduced, serum protein electrophoresis (SPEP) becomes the tool of choice. The test separates blood proteins by electrical charge and size, producing a pattern of bands on a gel or a graph with distinct peaks. A normal pattern shows a large albumin peak and smaller peaks for different globulin fractions. An abnormal spike — a monoclonal “M-spike” — indicates that a single clone of immune cells is churning out one specific antibody. This finding is a hallmark of plasma cell disorders, ranging from the relatively benign monoclonal gammopathy of undetermined significance (MGUS) to multiple myeloma.17PubMed Central. Does my patient with a serum monoclonal spike have multiple myeloma? The size of the M-spike is tracked over time using serial SPEP measurements to monitor disease progression or response to treatment.18PubMed Central. Long-Term Biological Variation of Serum Protein Electrophoresis M-Spike, Urine M-Spike, and Monoclonal Serum Free Light Chain Quantification: Implications for Monitoring Monoclonal Gammopathies
Identifying Specific Proteins with Immunoassays
The methods described so far measure total protein or broad categories of protein. When you need to know whether a specific protein is present and how much of it there is, you need a method that can tell proteins apart by their molecular identity. Immunoassays do this by exploiting antibodies — molecules that bind to one target protein with high specificity, like a lock fitting a key.
ELISA (enzyme-linked immunosorbent assay) is the workhorse of specific protein detection. In a sandwich ELISA, a “capture” antibody is fixed to the bottom of a well in a plate. The sample is added, and if the target protein is present, it sticks to the capture antibody. A second “detection” antibody, linked to an enzyme, then binds to a different part of the target protein. Adding a substrate for the enzyme produces a color change whose intensity is proportional to the amount of target protein. ELISA can detect proteins at extremely low concentrations — in some applications, down to fractions of a nanogram per milliliter.19PubMed Central. A novel double-antibody sandwich ELISA based on monoclonal antibodies against the viral spike protein detects porcine deltacoronavirus infection It is used everywhere from pregnancy tests and COVID rapid antigen tests to research labs quantifying cytokines.20PubMed. Development of sandwich ELISA for IL-10 quantification in olive flounder (Paralichthys olivaceus) and its application to protein-level analysis during viral hemorrhagic septicemia virus (VHSV) infection
Western blotting takes a different approach. Instead of capturing proteins in a well, it first separates proteins by size using gel electrophoresis, then transfers (“blots”) them onto a membrane. The membrane is then probed with antibodies specific to the protein of interest. Because the separation step comes first, a Western blot tells you both whether the target protein is present and what its apparent molecular weight is, which helps distinguish the real target from other proteins that an antibody might weakly cross-react with.21PubMed Central. Western blot: technique, theory, and trouble shooting Western blots are slower and more labor-intensive than ELISA, but they remain standard for confirming protein identity in research and diagnostic settings — the classic example being confirmation of HIV antibodies.
Mass Spectrometry and Proteomics
When the goal shifts from “how much protein” or “is this one protein present” to “which of the thousands of proteins in this sample are there, and in what relative amounts,” mass spectrometry takes over. In a typical proteomics workflow, a complex protein mixture is first digested into short peptide fragments using an enzyme. Those fragments are separated, usually by liquid chromatography, and then fed into a mass spectrometer, which measures each peptide’s mass with extreme precision. Software matches the observed masses to databases of known protein sequences to identify what was in the original sample.22PubMed Central. Enhanced protein identification using graphite-modified MALDI plates for offline LC-MALDI-MS/MS bottom-up proteomics
This bottom-up proteomics approach can identify and quantify thousands of proteins from a single cell lysate, tissue sample, or blood draw. It has become indispensable in drug development, cancer biomarker discovery, and understanding disease mechanisms. The tradeoff is cost, complexity, and time. A mass spectrometry proteomics experiment requires expensive instrumentation, specialized sample preparation, and significant bioinformatics work to interpret the data. It is not a method anyone uses when a simple concentration measurement will do — it is what you reach for when the question is genuinely complex.
Picking the Right Test for the Job
The sheer number of protein testing methods exists because no single method answers every question. Choosing the right one depends on a few practical considerations:
- Total vs. specific: Do you need the total amount of protein in a sample, or do you need to know whether a particular protein is present? Colorimetric assays and nitrogen-based methods measure total protein. ELISA and Western blot identify individual proteins. Mass spectrometry can do both.
- Sample type: A food matrix, a purified lab sample, a urine specimen, and a blood draw each come with different contaminants and concentration ranges that favor different methods.
- Throughput: Dipsticks and Bradford assays give answers in minutes for dozens of samples at once. A 24-hour urine collection takes a day by design. Mass spectrometry can take days from sample prep to final data analysis.
- Sensitivity: If the protein is present at very low concentrations, you need a sensitive method like ELISA or mass spectrometry. If you have milligrams of protein in a test tube, a biuret assay works fine.
A common mistake in both research and clinical settings is treating any protein assay result as an absolute truth. Every colorimetric assay is calibrated against a reference protein, and every result is an estimate relative to that standard. Two labs using different assays and different standards on the same sample can get different numbers and both be “right” according to their own calibration. This is why clinical labs report reference ranges specific to their instruments and reagents, and why regulatory bodies for food testing specify which method must be used for labeling compliance.
Portable and Point-of-Care Protein Testing
Much of the innovation in protein testing right now is focused on making sophisticated measurements available outside traditional labs. Paper-based microfluidic devices — essentially advanced versions of the lateral-flow strips used in home pregnancy and COVID tests — are being engineered to detect and quantify specific proteins with greater sensitivity than current rapid tests.23PubMed. Sensitive Protein Detection and Quantification in Paper-Based Microfluidics for the Point of Care The appeal is enormous: imagine a diabetic patient monitoring urinary albumin at home with a test strip that gives a real number, not just a positive-or-negative color change, or a field veterinarian checking serum protein levels in livestock without shipping samples to a distant lab. Advances in new detection labels, smartphone-based readers, and biosensor design are pushing these applications closer to reality, though most remain in development rather than routine clinical use.