Reactive oxygen species (ROS) can be measured using fluorescent probes, electron paramagnetic resonance (EPR) spectroscopy, chemiluminescence, or indirect biomarker assays. No single method detects all ROS equally well, and every technique has trade-offs in specificity, sensitivity, and practicality. Choosing the right approach depends on which ROS you need to detect, whether you’re working with live cells or tissue samples, and how quantitative your results need to be.
Fluorescent Probes for Live Cells
Fluorescent probes are the most accessible starting point for measuring ROS in cultured cells. They work by reacting with reactive species to produce a fluorescent product you can detect with a plate reader, flow cytometer, or fluorescence microscope. The three most common probes each target different species, but all come with important caveats about what they actually measure.
DCFDA for General Oxidative Stress
DCFH-DA is the most widely used probe for detecting intracellular oxidative stress, and it’s often described as a hydrogen peroxide sensor. That description is misleading. DCFH does not directly react with hydrogen peroxide to form its fluorescent product (DCF). Instead, multiple one-electron-oxidizing species drive the reaction: hydroxyl radicals, hypochlorous acid, reactive nitrogen species, and peroxidase-mediated reactions all contribute to the signal. The probe is better understood as a general redox indicator than a hydrogen peroxide detector.
A more serious problem is signal amplification through redox cycling. When DCFH is oxidized, the intermediate radical it forms reacts rapidly with oxygen to generate superoxide, which then produces additional hydrogen peroxide. This creates a self-amplifying loop that inflates fluorescence intensity beyond what the original oxidative stress would produce. For these reasons, DCFH-DA is not suitable for measuring specific intracellular oxidants. It can be useful as a broad indicator that something has shifted in cellular redox balance, but the signal cannot be attributed to any particular species.
DHE for Superoxide
Dihydroethidium (DHE) is the standard probe for intracellular superoxide. When DHE reacts with superoxide, it forms a specific product called 2-hydroxyethidium that serves as a diagnostic marker. This product is not formed when DHE reacts with other oxidants like hydroxyl radicals, peroxynitrite, or hydrogen peroxide, which instead generate ethidium and dimeric products. The catch is that 2-hydroxyethidium and ethidium have overlapping fluorescence spectra, so simple fluorescence measurements can’t distinguish between them. To confirm you’re actually measuring superoxide-specific signal, you need HPLC separation of the reaction products rather than just reading total fluorescence.
MitoSOX Red for Mitochondrial Superoxide
MitoSOX Red is a DHE derivative engineered to accumulate specifically in mitochondria. It carries a positively charged phosphonium group that drives it across the mitochondrial membrane, where it concentrates as a function of membrane potential. Once inside, it reacts with superoxide and becomes fluorescent upon binding mitochondrial DNA.
Loading concentrations vary by cell type and detection method. For flow cytometry, 5 µM at 37 °C in the dark for 10 to 30 minutes is typical across most cell types. For confocal microscopy of neurons, concentrations as low as 0.1 to 0.2 µM for 15 to 30 minutes are used to minimize background. Keratinocytes are commonly loaded at 5 µM for 15 minutes. In all cases, cells should be protected from light during loading and washed before imaging to remove unincorporated dye.
Amplex Red for Hydrogen Peroxide
If you specifically need to quantify hydrogen peroxide, the Amplex Red/horseradish peroxidase (HRP) assay is the most practical option. The enzyme HRP reacts with hydrogen peroxide, then oxidizes Amplex Red to resorufin, a highly fluorescent product. The assay detects as little as 5 picomoles of hydrogen peroxide per reaction, with a lower limit of quantification around 30 picomoles. It’s linear up to 4 µM hydrogen peroxide, making it useful across a wide working range.
The assay requires both Amplex Red (typically at 25 µM final concentration) and HRP (0.1 units per reaction) in phosphate buffer at pH 7.7. If your experiment involves NADPH, be aware that it causes a slow, enzyme-independent oxidation of Amplex Red that produces background fluorescence. This background isn’t blocked by superoxide dismutase but is abolished by catalase, confirming it comes from trace hydrogen peroxide generation. Subtracting this NADPH-dependent background from your readings corrects for the artifact. Other common buffer components, including iron chelators and superoxide dismutase, do not interfere with the assay.
EPR Spectroscopy: The Gold Standard
Electron paramagnetic resonance (EPR) spectroscopy is the only technique that directly detects free radicals. It works by measuring transitions of unpaired electrons in an applied magnetic field, using microwave radiation at a fixed frequency. Because ROS are defined by having unpaired electrons, EPR provides the most direct and specific measurement available, with sensitivity down to about 1 µM concentrations.
The challenge is that most biologically relevant ROS are extremely short-lived. To capture them, researchers use spin trapping: small molecules called spin traps react with fleeting radicals to form longer-lived “spin adducts” that persist long enough for detection. The most commonly used spin traps are DMPO and DEPMPO, both pyrroline-based cyclic nitrones that react with hydroxyl radicals and superoxide to form distinct adducts with identifiable spectral signatures. Another common trap, PBN, is useful but produces adducts whose spectra are less dependent on the identity of the trapped radical, making species identification harder.
EPR stands out because it works in complex, non-homogeneous biological solutions where other spectroscopic techniques struggle. It can also monitor changes in the oxidation state of transition metals involved in ROS generation, giving additional mechanistic information. The downsides are cost, the need for specialized equipment and expertise, and relatively low throughput compared to plate-based fluorescence assays.
Chemiluminescence Detection
Chemiluminescence offers another way to detect ROS in real time, using chemical probes that emit light when oxidized. The two most common probes, lucigenin and luminol, detect different species in different cellular compartments. Lucigenin localizes to mitochondria and its light emission reflects intramitochondrial superoxide production by the electron transport chain. Luminol, by contrast, detects hydrogen peroxide released from mitochondria into the surrounding environment. This compartmental difference means the two probes give complementary rather than redundant information when used together on the same sample.
Indirect Biomarkers of Oxidative Damage
When you can’t measure ROS directly, or when you want to assess cumulative oxidative damage rather than instantaneous ROS levels, biomarkers of oxidation products in blood or urine are the alternative. These don’t measure ROS themselves but measure the molecular damage ROS leave behind on lipids, proteins, or DNA.
F2-Isoprostanes
F2-isoprostanes are the recommended biomarker for monitoring oxidative status over time and for comparing oxidative stress between individuals. They’re formed when ROS attack cell membrane lipids and can be measured in both blood and urine. Their reliability for both longitudinal tracking and cross-sectional comparisons makes them the preferred clinical biomarker.
8-oxodG for DNA Damage
Urinary 8-oxodG reflects oxidative damage to DNA that has been repaired and excreted. It’s a validated, sensitive marker in both animal models and humans, with reference values in healthy children ranging from about 4.6 to 27.2 ng/mg creatinine. A practical advantage is that 8-oxodG levels aren’t significantly affected by diet or by long-term urine storage at -20 °C. The limitation is that urinary levels depend partly on individual DNA repair capacity, which varies between people. This makes 8-oxodG reliable for tracking changes within the same person over time but less ideal for comparing between individuals.
Malondialdehyde
Malondialdehyde (MDA) is frequently measured as a lipid damage marker, typically using a reaction with thiobarbituric acid. Despite its popularity, MDA has serious reliability problems. It isn’t specific to ROS-induced lipid damage, and urinary levels are confounded by dietary MDA content. In human clinical models of oxidative stress, even improved HPLC-based MDA assays failed to detect changes that were clearly present in animal models. Plasma reference values in healthy adults range roughly from 0.36 to 1.24 µmol/L, with men trending slightly higher than women and levels increasing with age. Given its lack of specificity and poor human validation, MDA is not recommended as a standalone biomarker of oxidative status.
Electrochemical Sensors for Real-Time Monitoring
Miniaturized electrochemical sensors represent a newer approach, borrowing design principles from continuous glucose monitors. These sensors can measure hydrogen peroxide and overall oxidation-reduction potential in real time, in small-scale biological environments. Early applications have focused on monitoring reactive species generated during cold atmospheric plasma therapy, where real-time feedback allows clinicians to adjust treatment doses on the fly. While not yet standard laboratory tools, electrochemical sensors point toward a future where ROS measurement moves from batch assays to continuous, point-of-care monitoring.
Choosing the Right Method
Your choice depends on what question you’re asking. If you need to identify a specific ROS species with high confidence, EPR with appropriate spin traps is the most definitive approach. If you’re screening for changes in general oxidative stress in live cells, DCFH-DA gives a quick readout, as long as you interpret it as a broad redox indicator and not a hydrogen peroxide measurement. For mitochondrial superoxide specifically, MitoSOX Red is the standard. For quantifying hydrogen peroxide with high sensitivity, the Amplex Red/HRP assay is hard to beat.
For clinical or epidemiological studies where you need to assess oxidative stress from a blood or urine sample, F2-isoprostanes are the strongest choice for between-person comparisons, while 8-oxodG works well for tracking the same individual over time. Whichever method you use, combining at least two complementary approaches strengthens your conclusions, since every ROS detection method has blind spots that a second technique can cover.