A dot plot with two population blobs and a polygon drawn around one of them is meaningless without knowing what happened to the cells before that plot: what was excluded as debris, what was excluded as a doublet, what was excluded as dead, and what the axes are actually measuring. Flow cytometry is a single-cell measurement technique — it passes cells one at a time through a laser and records how each one scatters light and fluoresces — and every plot downstream is a record of a filtering decision, not a raw picture of the sample. This guide walks through how the instrument physically works, what forward and side scatter actually report, how to read histograms, dot plots, and contour plots, and how to work through a standard gating sequence from raw events to the population you actually care about.
What Flow Cytometry Actually Measures
A flow cytometer suspends cells (or other particles — beads, nuclei, bacteria) in fluid and forces them, one at a time, through a narrow interrogation point where they cross one or more focused laser beams. As each cell crosses the beam, it scatters laser light in characteristic ways depending on its size and internal complexity, and any fluorescent molecules bound to or expressed by the cell — antibody-conjugated dyes, fluorescent proteins, viability dyes — are excited and emit light at a longer wavelength. Detectors positioned around the interrogation point convert that scattered and emitted light into electrical signals, and the instrument’s software records a full set of these measurements — one row of data — for every single cell that passes through, typically thousands to tens of thousands of cells per second. The output isn’t an image; it’s a table where every row is one cell and every column is one measured parameter (a scatter channel or a fluorescence channel), and the plots used to read the data are just different ways of visualizing that table.
The Three Systems: Fluidics, Optics, and Electronics
Fluidics: Hydrodynamic Focusing
The sample doesn’t simply drip past the laser — it has to be delivered one cell at a time, in single file, or the scatter and fluorescence readings from adjacent cells would blur together. Cytometers achieve this with hydrodynamic focusing: the cell suspension (the sample core) is injected into the center of a much faster-flowing stream of cell-free sheath fluid. Because the sheath fluid moves faster and surrounds the sample core symmetrically, it constricts the core stream down to a narrow thread only slightly wider than a single cell, lining cells up single file at the laser interrogation point. The ratio of sheath flow rate to sample flow rate controls both how tightly cells are focused (higher sheath flow narrows the core more, improving measurement precision) and how fast cells pass through the system (the acquisition rate).
Optics: Lasers and the Interrogation Point
At the interrogation point, one or more lasers of fixed wavelengths intersect the sample stream. Common laser lines on a standard configuration include violet (~405 nm), blue (~488 nm), yellow-green (~561 nm), and red (~633–640 nm); which lasers an instrument has determines which fluorophores it can excite well. As a cell crosses a laser beam, light is scattered in multiple directions and, if the cell carries fluorescent labels excitable by that laser’s wavelength, those fluorophores emit light of their own. A series of dichroic mirrors and bandpass optical filters splits this mixed light by wavelength and routes each slice to a dedicated detector, so that (for example) green emission around 519 nm and orange emission around 578 nm are measured on separate channels even though both fluorophores may have been excited by the same 488 nm laser.
Electronics: Detectors and Signal Digitization
Scattered and emitted light is weak, so cytometers use photomultiplier tubes (PMTs) — or, on newer instruments, avalanche photodiodes — to amplify the signal before it’s digitized. Each detector produces a voltage pulse as a cell crosses the laser; the instrument’s electronics characterize that pulse in a few standard ways — height (the peak voltage), area (the integrated signal under the pulse), and width (how long the pulse lasted) — and it’s the relationship between these pulse characteristics, not just their raw magnitude, that makes doublet discrimination possible (see the gating section below). The digitized values for every parameter, for every cell, are what get written to the list-mode data file (commonly an .fcs file) that gating software reads.
Forward Scatter and Side Scatter: What They Physically Report
Two scatter measurements are on essentially every flow experiment and form the backbone of the first gating step:
- Forward scatter (FSC) is light scattered at a narrow angle roughly in line with the laser beam. It correlates primarily with cell size — larger cells scatter more light forward. It’s a rough, relative measure, not a calibrated size in microns, but it’s reliable enough to separate cells from smaller debris and to distinguish broad size classes (small lymphocytes from larger monocytes, for example).
- Side scatter (SSC) is light scattered at roughly 90° to the laser beam. It correlates with internal complexity — granularity, membrane folding, nuclear-to-cytoplasm ratio, and organelle content. A granulocyte with dense cytoplasmic granules produces much higher SSC than a lymphocyte of similar size.
Plotted against each other, FSC and SSC produce the familiar first dot plot of a whole-blood or PBMC (peripheral blood mononuclear cell) sample, where lymphocytes, monocytes, and granulocytes separate into distinguishable clusters purely on the basis of size and granularity, before any fluorescent staining is even considered.
Fluorophores and Laser Lines
Choosing a fluorophore panel means matching each dye’s excitation peak to a laser line the instrument actually has, and its emission peak to a detector the instrument has fitted with the right filter — and making sure that dyes excited by the same laser don’t emit at wavelengths too close together to separate cleanly. The table below shows commonly used fluorophores and the laser line each is typically paired with; exact excitation/emission maxima vary slightly by manufacturer and dye formulation, so treat these as typical ranges rather than fixed values, and always confirm against your instrument’s configuration and the dye manufacturer’s spec sheet before designing a panel.
| Fluorophore | Typical laser line | Typical emission peak | Common use |
|---|---|---|---|
| FITC / Alexa Fluor 488 | Blue (488 nm) | ~519 nm (green) | Surface marker antibody conjugate |
| PE (phycoerythrin) | Blue (488 nm) or yellow-green (561 nm) | ~578 nm (orange) | Surface marker antibody conjugate |
| PerCP / PerCP-Cy5.5 | Blue (488 nm) | ~677–695 nm (red) | Surface marker antibody conjugate |
| APC (allophycocyanin) | Red (633–640 nm) | ~660 nm (far red) | Surface marker antibody conjugate |
| Pacific Blue / eFluor 450 | Violet (405 nm) | ~450 nm (blue) | Surface marker antibody conjugate |
| DAPI | Violet (405 nm) | ~461 nm (blue) | DNA content / dead-cell exclusion (compromised membranes only) |
| 7-AAD | Blue (488 nm) | ~647 nm (red) | Viability dye (excluded by intact membranes) |
| Propidium iodide | Blue (488 nm) or yellow-green (561 nm) | ~617 nm (orange-red) | Viability dye / cell-cycle DNA content |
Compensation and Spectral Overlap
Fluorophores don’t emit at a single wavelength — each has an emission spectrum, a range of wavelengths with a peak somewhere in the middle, and those spectra overlap. FITC, for instance, emits some light in the wavelength range typically used to detect PE, even though its peak is elsewhere. That means the detector assigned to the “PE channel” picks up a real signal from FITC-positive cells even if they aren’t stained with any PE at all — a false-positive signal that isn’t a mistake in staining, it’s a mathematical consequence of how light and filters work. Compensation is the mathematical correction applied after acquisition (or configured before it) that subtracts the calculated spillover of each fluorophore into every other fluorophore’s channel, using single-stain compensation controls (a sample stained with only one fluorophore at a time, run for every fluorophore in the panel) to measure exactly how much spillover exists.
On newer spectral cytometers, which measure the full emission spectrum of each cell across many detectors rather than one detector per fluorophore, the same underlying problem is handled with spectral unmixing instead of traditional compensation — the principle (mathematically separating each dye’s real contribution from another dye’s spillover) is the same, but the method and the reference controls used differ. Under-compensated or over-compensated data is one of the most common causes of a dot plot that looks wrong — populations that appear to smear diagonally, or a population that appears falsely double-positive — and it’s a data-processing artifact to check before concluding a biological result is real.
Controls: Unstained, FMO, and Isotype — What Each One Actually Tells You
These three controls are often confused with each other because all three involve “a tube that isn’t the full stain,” but they answer different questions and aren’t interchangeable:
- Unstained control: cells with no antibody or dye added at all. It establishes the cells’ natural autofluorescence — the baseline signal every detector picks up from the cells themselves, before any stain is added. It’s the starting reference point for setting initial voltages, not a gating boundary for a specific marker.
- Fluorescence-minus-one (FMO) control: the full staining panel, with every fluorophore included except one. It shows how much spillover from all the other dyes in the panel lands in that one channel, which is exactly the information needed to set an accurate positive/negative gate boundary for that specific marker in that specific panel. Because spillover depends on the whole panel, an FMO has to be run for every panel it’s used with — it isn’t reusable across different staining combinations.
- Isotype control: an antibody of the same isotype, species, and conjugate as the specific antibody being tested, but with no relevant specificity for the target antigen. It’s meant to estimate background staining caused by non-specific antibody binding (such as Fc receptor binding) rather than spillover. Isotype controls have fallen out of favor for setting gates in modern multicolor panels precisely because they don’t account for spillover the way an FMO does — a well-titrated FMO control is now the standard for gate-setting in most panels, with isotype controls reserved for specifically questioning whether a signal reflects true antigen binding versus non-specific antibody binding.
Using an isotype control where an FMO is needed (or vice versa) is a common source of a gate that’s set in the wrong place — know which question you’re actually trying to answer before choosing the control.
Reading the Plots: Histograms, Dot Plots, and Contour Plots
- Histogram: a single parameter on the x-axis (a fluorescence or scatter channel) against event count on the y-axis. Best for comparing the distribution of one marker across two or more samples or conditions overlaid on the same axes — for example, comparing a stained sample’s peak position and spread against an FMO control to see whether a population is genuinely shifted.
- Dot plot (scatter plot): two parameters plotted against each other, with one dot per cell (or, on dense datasets, dots that increasingly overlap). Best for identifying distinct populations defined by two markers simultaneously — a CD4 vs. CD8 dot plot, for instance, or the FSC vs. SSC plot used for the first gating step.
- Contour plot (or density plot): the same two-parameter comparison as a dot plot, but rendered as contour lines or a color density gradient instead of individual dots, so that overlapping populations in a dense dataset remain visually interpretable instead of collapsing into a solid mass of overlapping dots. Contour and density plots are generally easier to read than dot plots once event counts get into the tens of thousands.
All three are views of exactly the same underlying per-cell data table — switching from one plot type to another doesn’t change what was measured, only how it’s displayed.
The Standard Gating Sequence, Step by Step
“Gating” means drawing a boundary around a population of interest on a plot and restricting all downstream analysis to only the cells inside that boundary. A typical analysis works through gates in this order, each one narrowing the dataset before the next question is asked:
- Debris exclusion. On the initial FSC vs. SSC dot plot, a low-FSC, low-SSC cluster near the origin represents debris, dead cell fragments, and small particles rather than intact cells of interest. This population is excluded first with a gate drawn around the higher-FSC cluster of intact cells, before any other question is asked.
- Singlet discrimination (doublet exclusion). Two cells stuck together and passing through the laser as one event will register a similar peak height to a single larger cell but a longer pulse duration, since it takes longer for two stuck-together cells to cross the beam than one. Plotting FSC-Height against FSC-Area (or FSC-Area against FSC-Width, depending on the instrument) shows true single cells falling on a tight diagonal line, while doublets and clumps fall off that line with a disproportionately higher area or width relative to their height. A gate drawn around the diagonal singlet population removes doublets, which would otherwise appear as false double-positive events in later gates.
- Live/dead discrimination. Using a viability dye (7-AAD, propidium iodide, or a fixable amine-reactive viability dye, depending on whether the protocol includes fixation) gated on the singlet population, dead and dying cells — whose compromised membranes let viability dyes in, or whose amine-reactive dyes bind more total protein through a permeabilized membrane — are excluded. Dead cells bind antibodies non-specifically at a much higher rate than live cells, so skipping this gate is a common cause of spurious “positive” populations downstream.
- Population of interest. Only now, on the live, single, non-debris cells, are the actual experimental markers gated — for example, gating CD3+ T cells, then CD4 vs. CD8 within that T cell gate, then an activation or cytokine marker within the CD4 or CD8 subset. Each successive gate is drawn on the population defined by the gate before it, which is why gating strategies are usually diagrammed as a tree or hierarchy rather than a flat list of plots.
Getting the order right matters: gating on a marker of interest before excluding debris, doublets, and dead cells routinely inflates or distorts the apparent size and phenotype of the population being studied, because non-viable and non-singlet events haven’t been removed from the denominator yet.
Common Reading Mistakes
- Treating FSC as a precise, calibrated size measurement. It’s a relative, instrument-specific value useful for comparison within a run, not a micron measurement without calibration beads.
- Setting a positive/negative gate by eye against an unstained control instead of an FMO. Autofluorescence alone underestimates where spillover from other dyes in the panel will actually land.
- Skipping doublet discrimination on a panel that includes DNA content or cell-cycle analysis, where doublets are especially likely to be misread as a distinct ploidy population.
- Over-interpreting a small percentage difference between samples without checking equivalent event counts were collected in the relevant gate — a rare population’s percentage becomes statistically unstable at low event counts.
Frequently Asked Questions
What is the difference between an FMO control and an isotype control?
An FMO control (the full panel minus one fluorophore) measures spillover from the rest of the panel into that channel and is the standard way to set an accurate gate boundary. An isotype control (a non-specific antibody of matching isotype and conjugate) estimates non-specific antibody binding, not spillover. They answer different questions and aren’t substitutes for each other — see the controls section above.
What do FSC and SSC stand for, and what do they measure?
Forward scatter (FSC) and side scatter (SSC). FSC correlates with a cell’s relative size; SSC correlates with its internal granularity and complexity. Together they’re typically the first plot in a gating strategy, used to separate intact cells from debris and to distinguish broad cell types by size and granularity alone.
Why do I need to gate on singlets before anything else?
Two or more cells passing through the laser stuck together register as one event with an inflated pulse area or width relative to its height. Left ungated, these doublets can appear as false double-positive populations in downstream marker gates. Comparing pulse height to pulse area or width on a scatter channel isolates true single-cell events before any marker is analyzed.
What’s the difference between a dot plot and a histogram?
A histogram shows the distribution of one parameter (event count vs. signal intensity), which is ideal for comparing a single marker’s distribution across samples or against a control. A dot plot shows two parameters against each other, one dot per cell, which is ideal for identifying populations defined by two markers at once. Both are drawn from the same underlying per-cell data.
Why does my dot plot show a diagonal smear between two populations that should be separate?
This is the classic visual signature of a compensation problem — either under- or over-compensated spillover between two fluorophores whose emission spectra overlap. Check the single-stain compensation controls for the fluorophores involved before concluding the smear reflects a real biological intermediate population.
Flow cytometry data is downstream of two other common lab techniques worth understanding alongside it: cells are typically prepared from a live culture (see the cell culture basics guide), and antibody-based detection principles — specificity, titration, and validation — carry over directly from methods like western blotting and ELISA, which use many of the same antibodies against the same targets in a different readout format. For broader lab operations and equipment context, see the Laboratory Operations hub.
