How Much Template DNA or cDNA to Use Per qPCR Reaction

For most qPCR reactions, you want 1–100 ng of cDNA (measured as the original RNA input into the RT reaction) or 10–100 ng of genomic DNA per well. The sweet spot for gene expression work is typically 10–25 ng RNA-equivalent of cDNA, which lands most moderately expressed genes (think GAPDH, ACTB, housekeepers) around Ct 15–20 and your gene of interest somewhere in the Ct 20–30 range. That gives you enough dynamic range to quantify without running into inhibition problems or burning through your sample.

If you're working with low-abundance targets — a cytokine in unstimulated cells, a transcription factor expressed at a few copies per cell — you can push to 50–100 ng input. If you're quantifying something abundant like 18S rRNA, you might need to dilute down to 1–5 ng to keep the Ct above 10, where most instruments start losing linearity. The short version: there's no single correct amount, but there is a range that keeps you out of trouble, and going above it causes more problems than going below it.

Why More Template Isn't Always Better

The instinct to load more template "just to be safe" is one of the most common mistakes I see. At high template concentrations, you run into PCR inhibition — and not from some exotic contaminant, but from the sample itself. Excess cDNA carries over RT buffer components (DTT, random hexamers, dNTPs, reverse transcriptase) that can interfere with your polymerase. Genomic DNA preps bring salts, EDTA, ethanol carryover, or residual phenol depending on your extraction method. The result is a Ct value that's higher than expected, which you might misinterpret as low expression when it's actually inhibition suppressing your reaction.

Here's how this plays out in practice. Say you run a dilution series of your cDNA: neat, 1:5, 1:25, 1:125. If everything is clean, each 5-fold dilution should shift the Ct by about 2.32 cycles (log₂(5) = 2.32). If your neat sample gives a Ct of 22 but the 1:5 dilution gives a Ct of 23.5 — a shift of only 1.5 cycles instead of 2.3 — your concentrated sample is inhibited. The 1:5 or 1:25 dilution is actually giving you more reliable data than the neat sample with "more" template in it.

A few specific scenarios where excess template causes trouble:

• SYBR Green assays: Too much template increases the chance of nonspecific amplification, which shows up as melt curve shoulders or secondary peaks. You're giving off-target priming events enough substrate to compete.
• Multiplex TaqMan: Overloading template can cause one high-abundance target to consume shared reagents (dNTPs, polymerase) early, suppressing amplification of the lower-abundance target.
• High-GC targets: Excess template exacerbates secondary structure problems. Diluting the template sometimes fixes amplification failures that no amount of DMSO or betaine will solve.

Recommended Input Amounts by Application

These aren't arbitrary — they're based on what consistently gives Ct values in the reliable quantification range (Ct 15–32) across the most common experimental setups.

Gene expression (RT-qPCR with cDNA):

• Standard targets (GAPDH, B2M, HPRT1, most signaling genes): 10–25 ng RNA-equivalent per reaction
• Low-abundance targets (cytokines, receptors, transcription factors in resting cells): 25–100 ng
• Very high-abundance targets (18S, 28S rRNA): 1–5 ng, or use a 1:10 to 1:50 dilution of the same cDNA you use for other genes

Genomic DNA (copy number, ChIP-qPCR, genotyping):

• Standard targets: 10–100 ng gDNA per reaction
• ChIP-qPCR: typically 1–5 µL of immunoprecipitated DNA (concentration is usually too low to measure accurately; just use a fixed volume)
• Rare targets or cell-free DNA: you may need to go higher (up to 200 ng), but validate with a dilution series first

Plasmid standards:

• For standard curves, the template amount is defined by copy number, not mass. A typical range is 10¹ to 10⁷ or 10⁸ copies per reaction. At the high end, 10⁸ copies of a 5 kb plasmid is about 0.5 ng — so mass-wise, plasmid standards are extremely dilute. The main risk here is contamination from aerosolized high-copy tubes, not inhibition.

Reaction volume matters too. These amounts assume a 20 µL reaction. If you're running 10 µL reactions (common on 384-well setups on QuantStudio 5 or 7), halve the input. If you're running 25 µL reactions on an older CFX96 protocol, you can scale up slightly, but the concentration in the reaction — not the total mass — is what matters for inhibition.

How to Figure Out the Right Amount for Your System

The honest answer is: run a small pilot experiment. It takes one plate and an hour of hands-on time, and it saves you from weeks of confusing data.

Step 1: Make a dilution series of your cDNA. Take a representative sample (not your precious treated condition — use a control or a pooled sample). Make four to five serial dilutions: 1:1 (neat), 1:5, 1:25, 1:125, 1:625 of your cDNA. Run each in triplicate against one reference gene and one GOI.

Step 2: Plot Ct versus log(dilution factor). Calculate the efficiency from the slope: E = 10^(−1/slope) − 1. You want this between 90% and 110%, corresponding to a slope between −3.6 and −3.1. If efficiency is below 90% at the concentrated end but improves with dilution, you have inhibition. Use the dilution where efficiency normalizes.

Step 3: Check your Ct values. Your reference gene should land between Ct 15 and 25 at your chosen dilution. Your GOI should be at Ct 32 or lower. Above Ct 32–33, stochastic variation dominates — you'll see replicate CVs > 1.0 Ct, and your quantification becomes unreliable.

Step 4: Check replicate consistency. At your chosen input amount, technical replicate standard deviation should be < 0.5 Ct, ideally < 0.3 Ct. If replicates are noisy at a particular dilution, it's usually because you're at the extreme ends — either inhibited (too concentrated) or stochastic (too dilute).

This whole experiment is essentially a validation of your RT-qPCR workflow, and it's worth doing once per sample type. Switching from HeLa cells to primary T cells? Different tissue? Different RNA extraction kit? Run the pilot again. The carryover profile changes with every extraction method, and what works for column-purified RNA from an RNeasy kit may not work for TRIzol-extracted RNA from adipose tissue (which is notoriously dirty).

Common Edge Cases and Fixes

Your Ct values are in the high 30s or undetectable, even with maximum template input. Before you assume your gene isn't expressed, check: Did the RT reaction work? Run ACTB or GAPDH on the same cDNA. If the housekeepers also fail, the problem is upstream — degraded RNA, failed reverse transcription, or an inhibitor that's killing everything. If housekeepers are fine but your GOI is absent, then yes, it may genuinely not be expressed in your sample, or your primers need redesigning.

NTC (no-template control) shows amplification at Ct 36–39 in a SYBR Green assay. This is usually primer dimer, not contamination — check the melt curve. If it's a distinct lower-Tm peak (typically 5–10°C below your target), it's dimers. Your actual samples at Ct 20–28 are fine; the dimers only show up when there's no real template to outcompete them. If your GOI Ct values are above 33, though, primer dimers become a real quantification problem, and you may need to redesign primers or increase template.

You're comparing tissues with very different RNA quality or composition. Brain, liver, and muscle have wildly different transcriptomes and total RNA profiles. A fixed mass input (say, 20 ng) from each tissue doesn't mean you're loading equivalent amounts of mRNA — rRNA and tRNA content varies. This is one reason why reference gene validation (using tools like geNorm from Vandesompele et al., 2002, or NormFinder from Andersen et al., 2004) matters so much. It's also why you should use the same cDNA dilution strategy for every tissue and verify efficiency independently in each.

You're using a one-step RT-qPCR kit. With kits like Luna One-Step (NEB) or SuperScript III One-Step (Invitrogen), you're adding RNA directly to the qPCR reaction. Input recommendations are typically 1–100 ng total RNA per reaction, with 10–50 ng being the sweet spot. The same inhibition principles apply, but you also need to worry about RNA secondary structure at high inputs, which can impede the RT step.

A Practical Starting Point

If you just want a number to start with and optimize later: use 20 ng RNA-equivalent of cDNA per 20 µL reaction. This works for the vast majority of gene expression studies using PowerUp SYBR Green, Luna Universal, or iTaq Universal on a CFX96 or QuantStudio 3/5. It keeps housekeepers around Ct 18–22 and most GOIs comfortably within the quantifiable range.

If you're running standard curves or efficiency checks on your dilution series data, VoilaPCR calculates amplification efficiency, flags inhibition, and checks replicate consistency automatically — so you can spend your time on the biology instead of wrestling with spreadsheets.