Skip to main content
v2026.11,986 entries · CC-BY 4.0
LAC HealthSponsored

Shop medical supplies

  • FDA registered
  • ISO 13485
  • BBB Accredited A+

Shop at LAC Health

Fluorescence Correlation Spectroscopy (FCS): Principle, Instrument and Limits

Fluorescence correlation spectroscopy explained: how fluctuations give concentration and diffusion, the FCCS variant, the pM-nM window and error sources.

Written and maintained by CASRAI Editorial Board

Published

On this page

Fluorescence correlation spectroscopy (FCS) is a fluorescence method that measures how the fluorescence signal from a very small observation volume fluctuates over time. Because labelled molecules wander in and out of that volume, the time correlation of the fluctuations gives the mean number of molecules in the volume (a concentration) and how long they take to cross it (a diffusion time, hence a diffusion coefficient and hydrodynamic radius), rather than a brightness or a bulk spectrum.

Definition and scope

FCS is a statistical analysis, by time correlation, of stationary fluctuations of fluorescence intensity. A 2021 review in Frontiers in Physics describes the signal as the fluorescence of labelled molecules diffusing in and out of a tiny observation volume; the small volume is what makes the fluctuations large enough to analyse and keeps measurement times short. A 2011 review in Applied Spectroscopy by Tian, Martinez and Pappas puts the contrast plainly: FCS does not rely on the intensity of emission but on spontaneous fluctuations, and those fluctuations can come from diffusion as well as from chemical reactions, quenching and conformational change.

That makes FCS a different measurement from the bulk method described in fluorescence spectroscopy, where a spectrofluorometer reports the average emission of the whole cuvette. It also differs from FRAP, which bleaches a region on purpose and watches recovery; FCS perturbs nothing and measures equilibrium fluctuations in a sample that is, on average, not changing. Both are mobility measurements, and the choice between them is covered below.

The method is a statistical average over many molecules passing through the volume, not tracking of individual molecules, even though the observation volume is small enough that the number of molecules in it is only a handful at a time. The 2022 principles review in Biophysics Reports lists what auto-correlation and cross-correlation FCS can quantify together: molar concentration, diffusion coefficient and hydrodynamic radius, homo- or hetero-interaction, and fluorescence brightness.

How FCS works in practice

The observation volume and the instrument

A typical instrument is a confocal microscope. According to Yu and colleagues (2021), the confocal volume is about 0.5 fL, with a pinhole in the detection path in front of an avalanche photodiode (APD). Tian and colleagues give the confocal detection volume as usually about 1 fL, and note that total internal reflection (evanescent) excitation shrinks it to the attolitre range. The two figures differ because the volume depends on the objective, wavelength and pinhole; neither is a universal constant, so a laboratory should measure its own volume rather than copy either number.

Yu and colleagues also explain that a pulsed laser with time-correlated single-photon counting (TCSPC) is preferred to a continuous-wave laser in demanding work, because it helps reject background and allows pulsed interleaved excitation. The correlation itself is computed either by a hardware correlator or in software, using the multi-tau algorithm, which spaces lag times logarithmically so a single curve spans many decades of time.

Reading the autocorrelation curve

The curve, G(τ), is the self-similarity of the signal after a lag time τ. Two features carry most of the information. The amplitude, G(0), is inversely proportional to the average number N of molecules in the observation volume: the fewer the molecules, the larger the relative fluctuations. The width or decay time, τD, is the mean time a molecule stays in the volume because of diffusion (Jovicic and colleagues, 2024); it follows τD = r0² / (4D) for a lateral volume radius r0, so a slower molecule gives a curve that decays later. The hydrodynamic radius then follows from the diffusion coefficient through the Stokes–Einstein relation.

Fitting needs a model. For free three-dimensional diffusion in solution, Yu and colleagues describe a fit with three free parameters: the amplitude, the diffusion time and a structure parameter describing the ratio of the axial to the lateral dimension of the volume. Tian and colleagues also describe a two-dimensional model for molecules diffusing in a membrane and a flow model for active transport, and list the quantities a fit returns: diffusion time, mean molecule number, diffusion coefficient and molecular brightness. The data table below sets out each fitted quantity beside the errors that corrupt it.

Concentration window

FCS works only in a window of low concentration. Yu and colleagues give a measurable range of roughly 1 pM to about 100 nM; Tian and colleagues cite about 10 pM to 10 nM for microfluidic applications, depending on the instrument. The upper limit has a simple cause: with more molecules in the volume, each arrival or departure changes the signal by a smaller fraction, so the fluctuation amplitude shrinks into the noise. The lower limit comes from how rarely a molecule enters the volume, which makes the measurement slow and noisy. Many biological samples sit above the window in their native state, which is why live-cell FCS usually works with low-expressed or dilute labelled species.

The variants below each solve one limitation of single-colour FCS. Every row traces to the sources named in the last column.

Variant What changes What it adds Source
Single-colour FCS (autocorrelation) One dye, one detection channel Concentration, diffusion time, brightness Yu et al. 2021; Tian et al. 2011
Fluorescence cross-correlation spectroscopy (FCCS) Two spectrally distinct dyes, two channels A complex fluctuates in both channels, so binding can be read directly and a KD estimated from a binding curve Yu et al. 2021
Dual-focus FCS (2fFCS) Two known, shifted foci Suggested to reduce dependence on the volume shape; better for flow Jovicic et al. 2024; Tian et al. 2011
Total internal reflection FCS Evanescent-wave excitation Attolitre-scale detection volume near a surface Tian et al. 2011
Pulsed excitation with TCSPC Pulsed laser and photon timing instead of a continuous laser Background rejection; pulsed interleaved excitation Yu et al. 2021

FCCS is the variant researchers most often ask about. Tian and colleagues point out the reason it exists: standard FCS struggles when a binding event changes mass only slightly, as in an enzyme versus its complex, because a diffusion time depends only weakly on mass. Two colours detect the complex by co-movement instead. FCCS demands good spectral separation between the dyes and a precise overlap of the two observation volumes.

Where it matters for labs and instrument decisions

What FCS is used to measure

The reviews list a consistent set of applications. Yu and colleagues cite diffusion coefficients and hydrodynamic radius (for example, the change in nanoparticle size when a protein adsorbs), local concentration, interactions through FCCS such as receptor-ligand binding, viscosity, and rotational diffusion on the nanosecond to microsecond scale. Tian and colleagues describe microfluidic flow profiling, membrane receptor density, enzyme kinetics, and the cleavage of a caspase-3 probe in Jurkat cells. Jovicic and colleagues add free-solution studies, membrane dynamics, protein folding and amyloid-β aggregation. The Biophysics Reports review applies auto- and cross-correlation FCS to biomolecular liquid–liquid phase separation, in vitro and in live cells.

Calibration and quality checklist

The quantitative value of an FCS result depends on knowing the observation volume and keeping it stable. The three reviews give a coherent checklist, drawn together here from their separate statements:

  • Calibrate with a reference dye of known diffusion. Jovicic and colleagues used standard 10 nM solutions of Rhodamine 6G and Alexa 633 for instrumental calibration and glycerol solutions of known viscosity to check that measured diffusion behaved as expected. Tian and colleagues advise checking alignment daily with a dye of known diffusion coefficient.
  • Control the optical parameters. Jovicic and colleagues note that quantitative results depend on optical parameters such as refractive index and cover-slip thickness.
  • Keep excitation below saturation. The same review calls optical saturation the most notable cause of inaccuracy and irreproducibility in FCS results and suggests dual-focus FCS as a remedy.
  • Watch for photobleaching. Bleaching during the measurement lowers the apparent number of molecules and shortens apparent diffusion times (Yu et al. 2021). The photophysics are explained in fluorescence quenching and photobleaching, and the choice of dye in fluorophores.
  • Measure the background. Autofluorescence and detector dark counts add an uncorrelated signal that lowers the amplitude and so inflates the apparent molecule number (Yu et al. 2021).
  • Check dual-colour overlap. Spectral crosstalk produces false-positive cross-correlation, and imperfect overlap of the two volumes produces false negatives (Yu et al. 2021).

FCS or FRAP?

Both answer a question about how fast labelled molecules move, but they work at different scales and in different ways. The site’s FRAP guide describes FRAP as bleaching a small region with an intense laser pulse and imaging the recovery as unbleached molecules diffuse in; the recovery gives a half-time and diffusion coefficient and the plateau gives the mobile fraction. FCS instead needs a low concentration, bleaches as little as possible, and reads fluctuations at a point. In the sources read here, FCS is therefore the method for dilute, fast-moving species and for interactions seen by cross-correlation, while FRAP reports a mobile fraction that FCS does not directly provide. Neither measures an interaction on its own without a second experiment or a second colour.

Common misunderstandings

  • “A bright sample gives a better curve.” The opposite holds. Amplitude is inversely proportional to the number of molecules, so a concentrated sample washes out the fluctuations. The method is built for picomolar to nanomolar solutions.
  • “FCS tracks single molecules.” It reports an ensemble statistic of many brief passages through the volume. Yu and colleagues note that it is a statistical average, not single-molecule tracking.
  • “Aggregates are just large particles.” In samples that aggregate, a few bright aggregates distort the curve and the fit becomes unreliable (Yu et al. 2021). Amyloid-β aggregation studies treat this as a subject in its own right rather than a nuisance (Jovicic et al. 2024).
  • “The numbers transfer between laboratories.” They transfer only when the observation volume is calibrated, as above. Tian and colleagues also caution that results can mislead in different environments and that interpreting models are lacking near boundaries and in microchannels.
  • “A cross-correlation means binding.” It can mean crosstalk. Both dyes must be checked on single-label controls.

Sources

Data: Fitted FCS quantities and the errors that corrupt them

Compiled from four peer-reviewed reviews of fluorescence correlation spectroscopy: for each fitted quantity (amplitude, decay time, brightness) what it reports, and for each recognised error source what it does to the result. Each row is stated in at least one of the listed reviews; the table pairs the statements so a reader can see what to check before trusting a number.

Period covered: 2011/2024.

Fitted FCS quantities and the errors that corrupt them
ItemKindWhat it is or what it doesEffect on the resultSource
Amplitude G(0)Fitted parameterInversely proportional to the mean number N of molecules in the observation volume, so it reports local concentrationFalls toward noise as concentration rises above about 100 nMYu et al. 2021; Tian et al. 2011; Jovicic et al. 2024
Decay time τDFitted parameterMean time a molecule stays in the observation volume by diffusion; gives the diffusion coefficient D and, via Stokes-Einstein, the hydrodynamic radiusDepends on the calibrated volume size, refractive index and cover-slip thicknessJovicic et al. 2024; Yu et al. 2021
Molecular brightnessFitted parameterCount rate per molecule, returned by the fit alongside N, D and tau_DReduced by photobleaching and by optical saturationTian et al. 2011
PhotobleachingError sourceDye destroyed while crossing the volumeLower apparent N and shorter apparent diffusion timesYu et al. 2021
Background and dark countsError sourceUncorrelated signal from autofluorescence and detector countsLower curve amplitude, distorting NYu et al. 2021
Optical saturationError sourceExcitation power high enough that emission no longer scales with intensityNamed the most notable cause of inaccuracy and irreproducibility in FCS resultsJovicic et al. 2024
AggregatesError sourceA few bright, slow particles among many small onesDistorted curves; FCS is unreliable for aggregating samplesYu et al. 2021
Spectral crosstalk (FCCS)Error sourceDye emission leaking into the other channelFalse-positive cross-correlationYu et al. 2021
Imperfect volume overlap (FCCS)Error sourceTwo foci not coincidentFalse-negative cross-correlationYu et al. 2021; Tian et al. 2011

Follow CASRAI

Research-administration guidance, standards updates and independent tool reviews.

Ask CASRAI · full access

Ask about Fluorescence Correlation Spectroscopy (FCS): Principle, Instrument and Limits

Full access: large monthly usage, up to 10 source passages behind each answer, and MCP and API access for Claude Code, Codex and other CLIs. Cancel any time.

An AI assistant for research: funder rules, compliance, methods and publishing. It links the sources behind its answers and labels answers that come from a web search.

Answers draw on CASRAI's guides and dictionary plus the federal and funder documents we index: Federal Register, Grants.gov, Regulations.gov and UKRI.

Works on this site and inside Claude, Cursor and the AI tools you already use.

Everything CASRAI publishes — this page, the dictionary, the guides and the news — stays open to read, with no account and no card.

Referenced across the research world

University of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logoUniversity of Cambridge logoColumbia University logoCrossref logoUniversity of Edinburgh logoHarvard University logoUniversity of Oxford logoPrinceton University logoStanford School of Medicine logoUniversity College London logoORCID logo
  • University of Cambridge logo
  • Columbia University logo
  • Crossref logo
  • University of Edinburgh logo
  • Harvard University logo
  • University of Oxford logo
  • Princeton University logo
  • Stanford School of Medicine logo
  • University College London logo
  • ORCID logo

View CASRAI adoption →