Skip to content
BioSyn

Delta Delta Ct Fold-Change Calculator

The ΔΔCt (delta-delta-Ct) method turns four qPCR threshold values into a single fold-change number. You normalise the gene of interest to a housekeeping gene in both a test and a control sample (that gives two ΔCt values), subtract them to get ΔΔCt, and report fold change = 2^(−ΔΔCt). A ΔΔCt of −3 means an 8-fold increase, +3 means an 8-fold decrease, and 0 means no change. Enter the four mean Ct values below and the calculator returns ΔCt for each sample, ΔΔCt, the fold change, the log2 fold change and, if you supply replicate standard deviations, a fold-change confidence range.

Relative quantification asks a simple question: how much more (or less) of my gene of interest is in the treated sample than in the control? qPCR does not measure molecules directly — it measures the cycle at which fluorescence crosses a threshold, the Ct. Because amplification is (ideally) exponential, each extra cycle of Ct means the sample started with about half as much template, so differences in Ct translate into fold differences through a power of two. This page is part of the PCR & qPCR calculator set and implements the Livak 2^(−ΔΔCt) method, the standard way to make that translation.

There are two normalisation steps, and both matter. The first removes sample-to-sample variation in how much total RNA or cDNA you loaded: you subtract the reference (housekeeping) gene Ct from the target gene Ct within the same sample, giving ΔCt. The second compares conditions: you subtract the control ΔCt from the test ΔCt, giving ΔΔCt. Fold change is then 2^(−ΔΔCt). The minus sign in the exponent is the part people forget — Ct is inversely related to starting amount, so a lower Ct in the test sample means more template and a fold change greater than one.

Why the base 2? The method assumes each PCR cycle doubles the amplicon, i.e. 100% efficiency. Under that assumption a one-cycle Ct difference is exactly a two-fold difference in starting material, so 2^(−ΔΔCt) is the fold change. That is also why you should confirm efficiency before trusting the number: if your assay runs at 85% or 115%, the base is not 2 and the Livak result is biased. The qPCR efficiency calculator is the companion quality-control step, and when efficiencies are far from 100% or unequal between target and reference you should switch to the Pfaffl efficiency-corrected model.

The choice of reference gene is the other silent assumption. ΔCt normalisation only cancels loading error if the housekeeper is expressed at a constant level across your test and control conditions. A gene that itself responds to the treatment will absorb or invent a fold change. Validate reference stability for your specific experiment; common choices (GAPDH, ACTB, 18S) are not universally stable. Finally, the control sample you designate is the 1× baseline — fold change is always relative to it, never absolute, so state the calibrator whenever you report the result.

Mean Ct of the gene of interest in the treated/test sample.

Mean Ct of the housekeeping gene in the test sample.

Mean Ct of the gene of interest in the control/calibrator sample.

Mean Ct of the housekeeping gene in the control sample.

Replicate SD of the test target Ct. Supply all four SDs for an error range.

Replicate SD of the test reference Ct.

Replicate SD of the control target Ct.

Replicate SD of the control reference Ct.

Fold change (2^−ΔΔCt)8 ×
ΔCt (test)2 cycles
ΔCt (control)5 cycles
ΔΔCt-3 cycles
log2 fold change3
Directionup
Fold-change range—

How to use the delta delta ct calculator

Run your qPCR with the target (gene of interest) and a reference (housekeeping) gene, each in replicate, for both a test/treated sample and a control/calibrator sample. Confirm efficiency is close to 100% first with the qPCR efficiency calculator.

Average the technical replicates for each of the four reactions to get four mean Ct values.

Enter those four means: target and reference Ct for the test sample, and target and reference Ct for the control sample.

Optionally enter the standard deviation of each mean Ct (from your replicates). Supply all four and the calculator propagates them into a ΔΔCt SD and a fold-change range.

Press Calculate. The primary card is the fold change; the supporting cards show ΔCt (test), ΔCt (control), ΔΔCt, log2 fold change and the direction of regulation.

Interpret: fold change > 1 is up-regulation, < 1 is down-regulation, = 1 is no change. Check that the error range (if given) does not straddle 1 before calling a result significant.

Delta Delta Ct Fold-Change formula

ΔCt(test) = Ct(target, test) − Ct(reference, test)
ΔCt(control) = Ct(target, control) − Ct(reference, control)
ΔΔCt = ΔCt(test) − ΔCt(control)
Fold change = 2^(−ΔΔCt)
log2 FC = −ΔΔCt
s(ΔΔCt) = √(s_tT² + s_rT² + s_tC² + s_rC²) (optional, from replicate SDs)
The factor 2 assumes 100% amplification efficiency (each cycle doubles the amplicon). If efficiency E differs from 100%, replace 2 with (1 + E); if the target and reference efficiencies differ, use the Pfaffl model instead of Livak. The error range, when supplied, is symmetric in ΔΔCt but asymmetric in fold change because 2^x is exponential.

Variables and units

SymbolMeaningUnit
CtThreshold cycle of one reaction (mean of replicates)cycles
ΔCtReference-normalised Ct of one samplecycles
ΔΔCtDifference between test and control ΔCtcycles
Fold changeRelative expression, 2^(−ΔΔCt)×
log2 FCLog2 of the fold change, = −ΔΔCt—
s(ΔΔCt)Propagated SD of ΔΔCt from replicate SDscycles

How delta delta ct is calculated

  1. Normalise within the test sample: ΔCt(test) = Ct(target, test) − Ct(reference, test).
  2. Normalise within the control sample: ΔCt(control) = Ct(target, control) − Ct(reference, control).
  3. Compare conditions: ΔΔCt = ΔCt(test) − ΔCt(control).
  4. Convert to fold change: fold change = 2^(−ΔΔCt); log2 fold change = −ΔΔCt.
  5. Label the direction: up if log2 FC > 0, down if < 0, unchanged if 0.
  6. If all four replicate SDs are given, propagate s(ΔΔCt) = √(Σ s²) and report the fold-change range 2^(−(ΔΔCt ± s)).

Worked laboratory examples

Example 1 — An 8-fold induction

Given: test: target Ct 22, reference Ct 20 · control: target Ct 25, reference Ct 20

  1. ΔCt(test) = 22 − 20 = 2.
  2. ΔCt(control) = 25 − 20 = 5.
  3. ΔΔCt = 2 − 5 = −3.
  4. Fold change = 2^(−(−3)) = 2³ = 8.

Result: 8-fold up-regulation (log2 FC = +3)

Example 2 — A 4-fold repression

Given: test: target Ct 27, reference Ct 20 · control: target Ct 25, reference Ct 20

  1. ΔCt(test) = 27 − 20 = 7.
  2. ΔCt(control) = 25 − 20 = 5.
  3. ΔΔCt = 7 − 5 = 2.
  4. Fold change = 2^(−2) = 0.25.

Result: 0.25× — a 4-fold decrease (log2 FC = −2)

Example 3 — No change

Given: test: target 24, reference 20 · control: target 25, reference 21

  1. ΔCt(test) = 24 − 20 = 4; ΔCt(control) = 25 − 21 = 4.
  2. ΔΔCt = 4 − 4 = 0.
  3. Fold change = 2⁰ = 1.

Result: 1× — unchanged (log2 FC = 0)

Example 4 — Fold change with an error range

Given: the Example 1 Ct values, each mean Ct with a replicate SD of 0.2

  1. ΔΔCt = −3, so fold change = 8.
  2. s(ΔΔCt) = √(0.2² × 4) = √0.16 = 0.4.
  3. Lower = 2^(−(−3 + 0.4)) = 2^2.6 ≈ 6.06; Upper = 2^(−(−3 − 0.4)) = 2^3.4 ≈ 10.56.

Result: Fold change 8, range ≈ 6.1–10.6 — well clear of 1, so confidently up

Want the reasoning behind the numbers? Read How to Calculate Delta Delta Ct (2^-ΔΔCt).

How to interpret the result

Fold change is a ratio, not an amount. "8×" tells you the treated sample has eight times the normalised target expression of the control you designated — say which control, or the number is meaningless. A value below 1 is down-regulation; its reciprocal is the fold decrease (0.25× = a 4-fold drop).

log2 fold change is often the cleaner number to report and plot: +3 and −3 are symmetric 8-fold up and down, and 0 is no change. It is simply −ΔΔCt, so it carries the same assumptions.

The error range matters more than the point estimate. If the fold-change range straddles 1 (for example 0.7–1.4), the data do not distinguish regulation from noise, no matter how the central value looks. That is why entering replicate SDs is worth the effort — without them you get a single number with no sense of its reliability.

A surprising fold change usually traces back to an assumption that failed, not the arithmetic. Check efficiency first (the qPCR efficiency calculator tells you if the base should really be 2), then reference-gene stability, then whether the control was handled identically. The PCR master mix calculator helps keep reactions consistent across samples.

Assumptions

  • Amplification efficiency is close to 100% and approximately equal for the target and reference amplicons, so the base 2 is valid.
  • The reference (housekeeping) gene is stably expressed across the test and control conditions.
  • The four mean Ct values come from technical replicates run under the same threshold and conditions.
  • The designated control sample is the intended 1× calibrator for the comparison.
  • When an error range is requested, the four replicate SDs reflect independent measurement error.

Limitations

  • Not valid when efficiencies deviate markedly from 100% or differ between target and reference — use the Pfaffl efficiency-corrected model in that case.
  • A single ΔΔCt without replicate SDs gives no confidence range and cannot be judged against run-to-run noise.
  • Relative only: it never reports absolute copy number or concentration.
  • Garbage reference genes silently corrupt the result; the calculator cannot detect an unstable housekeeper.

Common mistakes

  • Forgetting the minus sign and reporting 2^(ΔΔCt) instead of 2^(−ΔΔCt), which inverts up- and down-regulation.
  • Skipping the efficiency check and applying the base-2 method to an assay running at 85% or 115%.
  • Using a reference gene that itself responds to the treatment, absorbing or inventing the fold change.
  • Averaging Ct across biological samples instead of within technical replicates, then treating the result as a single observation.
  • Calling a fold change significant when its confidence range still straddles 1.

FAQ

What does ΔΔCt stand for?

Delta-delta-Ct is the difference between two ΔCt values. Each ΔCt is a sample’s target Ct minus its reference Ct (reference-normalised). ΔΔCt is then ΔCt(test) minus ΔCt(control), and fold change = 2^(−ΔΔCt).

Why is fold change 2 to the power of negative ΔΔCt?

Because Ct is inversely related to starting amount: one cycle later means about half the template. With 100% efficiency each cycle is a factor of two, so the fold difference between test and control is 2^(−ΔΔCt). The negative sign converts the inverse Ct relationship into the correct direction.

What is a good ΔΔCt result?

There is no universal cutoff — it depends on the biology. What matters is that amplification efficiency is near 100%, the reference gene is stable, and the replicate-derived error range does not straddle 1. A confident result has a fold-change range clearly above or below 1.

When should I not use the Livak method?

When the target and reference amplification efficiencies differ from each other or from 100%. Measure efficiency first with the qPCR efficiency calculator; if it is off, switch to the Pfaffl efficiency-corrected model, which uses the measured efficiencies instead of assuming a base of 2.

Which sample is the control (calibrator)?

The condition you compare everything against — typically untreated, time-zero or wild-type. Fold change is always relative to it, so the same Ct data give a different fold change if you pick a different calibrator. Always state which sample you designated.

Do I need to enter the standard deviations?

They are optional, but without them you get only a point estimate and no confidence range. Enter the SD of each of the four mean Ct values from your technical replicates and the calculator returns a ΔΔCt SD and a fold-change interval, which is what lets you judge whether the change is real.

Scientific Sources

External links open in a new tab and carry rel="nofollow". We cite primary references (NIST, NIH/NCBI, peer-reviewed papers) and established manufacturer technical documentation (NEB, Thermo Fisher, Promega, IDT). Calculated results are estimates; your protocol and manufacturer guidance take precedence.

Last reviewed: 2026-09 · Method version: 1.0