explainer
Extension time and amplicon length
Match extension time to amplicon length, see why a short hold leaves partial products, and why a final extension cannot repair that miss.
- Author
- EVRINTH Editorial Team
- Published
- 8 October 2026
- Updated
- 8 October 2026
- Reading time
- 8 min

Extension time is how long you give the polymerase, at its working temperature, to copy from one primer to the other. Amplicon length is the distance it has to travel. Too short a hold leaves partial products that never become full templates. A final extension at the end of the programme does not repair that miss. This page explains the relationship and the limits of the usual planning figure. The rest of the cycle is in how polymerase chain reaction works.
The decision the clock is making
You are choosing a hold that is long enough for the full amplicon and not an invitation to copy every weak start in the tube. Endpoint PCR will then show presence and size if the primers and the template cooperate. It will not show how many copies you started with. If the band at the designed length is missing and a smear of shorter DNA appears only in the sample lanes, extension time is a prime suspect. If the no-template lane shows the same smear, you have a primer artefact and a longer hold will feed it.
Calculate the length from the primers, not from the gene's common name. The amplicon includes both primers. Flanking primers on a plasmid include the vector bases between them. A multiplex is timed for its longest member, or that member drops out while the short ones look healthy.
What the enzyme is doing during the hold
At the extension temperature the polymerase adds nucleotides to each annealed primer and walks toward the other end. For a strand to become a template in the next cycle, the walk has to reach the sequence where the opposite primer binds. A strand that stops halfway lacks that site. The next cycle can copy it only as far as it goes, so the full-length product never enters exponential growth. You may see nothing at the expected size, or a smear of partial lengths. Those partials are a timing result when a longer hold, on a fresh tube, replaces them with one band. How that gel is read is in agarose gel electrophoresis for DNA.
Processivity is how far the enzyme tends to go before it falls off. A longer hold helps an enzyme that is still attached or that can rebind. It does not turn a polymerase with poor processivity into a long-range enzyme. Products of many kilobases are often a different enzyme class, marketed for long amplification, not a standard Taq recipe with an hour-long hold. Very GC-rich stretches can stall the enzyme even when the clock looks generous. That stall is a template-structure problem as well as a timing problem.
The temperature of the hold belongs to the enzyme. Classical Taq-like reactions are often extended near 72 Celsius. Another polymerase may ask for a different set-point. Running a Taq time at the wrong temperature is not the same experiment as running it at 72.
The planning figure, and when it does not apply
For classical Taq-like enzymes, a widely used planning figure is about one minute per kilobase of amplicon, near 72 Celsius. A 400-base-pair product is then well under a minute. A 3-kilobase product is on the order of three minutes. Treat the figure as a start. Some Taq cards allow a shorter time per kilobase. Some reactions with difficult templates need longer than the figure. The band on the gel is the confirmation.
Engineered polymerases publish their own speeds, sometimes well under a minute per kilobase, sometimes with a minimum hold even for short products. Copy that card. Putting a fast enzyme's seconds-per-kilobase onto Taq leaves partial products. Putting the Taq minute-per-kilobase onto a fast enzyme mostly wastes time, and on a low-specificity pair it can give weak off-target starts long enough to finish. Match the enzyme in the tube, not the enzyme in the paper you learned from.
Short amplicons rarely fail because one minute was not enough. If a 150-base-pair product is missing, look at primers, annealing temperature and template before you lengthen extension. Long amplicons fail the other way: the programme was copied from a genotyping assay of a few hundred bases, and the new 4-kilobase product never had a chance. Rewrite the hold when the length changes. Do not keep a single programme for every primer pair in the freezer.
| Amplicon situation | Planning start | What a too-short hold does | What will not fix it |
|---|---|---|---|
| Classical Taq-like, ordinary length | About one minute per kilobase near 72 Celsius | Partials, or no full-length band | A final extension after every cycle was short |
| Engineered fast polymerase | The speed on that card | Same partials if you used the Taq minute by mistake | Assuming all "Taq" labels share one speed |
| Product of many kilobases | A long-range enzyme class, then its timing | A stalled smear even with a long Taq hold | Extending for an hour on an enzyme that falls off |
| Multiplex with mixed lengths | The longest product's requirement | The long band drops, short bands remain | Timing the average length |
| Very short product | The card's minimum hold | Usually nothing, because time was not the limit | Blaming extension before you check primers |
Why the final hold is a different job
Many programmes end with a few extra minutes at the extension temperature. That final extension gives strands that nearly finished in the last cycle time to reach the end. For Taq-like enzymes it also supports the non-templated A overhang some cloning methods use. It runs once. The exponential copies were made under the per-cycle hold. If that per-cycle hold was half of what the length required, the molecules that would have needed the missing minute were never full templates, and one last incubation cannot invent them.
Use a final extension for the reason the card states. Do not lengthen it to compensate for a stingy cycle. Lengthen the cycle extension, leave the final hold as the card suggests, and rerun. If the full-length band appears, the clock was the constraint. If it does not, the enzyme's processivity, the denaturation of a long or GC-rich template, or the primers are next. A longer final hold is the wrong next experiment.
The cycler can add or remove time you did not programme
The hold is time at temperature only if the block gets there when the clock starts. A slow ramp spends additional time in the extension range on the way up from annealing, so an old programme can succeed on a slow cycler and fail when it is copied onto a fast block that reaches 72 Celsius and then counts a shorter effective synthesis. The reverse also happens: a fast-cycler protocol moved onto a slow block spends so long ramping that the enzyme, which is not immortal at high temperature, loses activity before a long amplicon is done.
When you move a programme, re-time from the card for that enzyme and that instrument class. Confirm with one known template whose length you can see on a gel. Two displays of "72 Celsius, 60 seconds" are not the same liquid history.
A long template that never fully denatures looks like a failed extension. If a longer hold does nothing, try the denaturation the card allows before you add more minutes at 72 Celsius.
Safety and a clock that lost its place
Extension troubleshooting ends at a hot block and a stained gel. Use the ultraviolet shield, and follow the stain note. Long amplicons from a regulated template are still that template. This explanation is not a diagnostic timing chart.
A power cut mid-programme stops the extension clock and the record of how far the run got. Do not resume a plate and call the product fully extended. Start a new setup, especially for a long amplicon where a partial run is indistinguishable from the failure mode on this page. In a shared room, write the enzyme name next to the seconds per kilobase you actually used. The next person will otherwise apply your fast-enzyme timing to a Taq tube, or your Taq timing to a fast enzyme, and both will misread the gel.
What to send with an enzyme question
Send the amplicon length in base pairs, the enzyme class you have been using, the extension hold you programmed, and whether the gel showed a full band, a partial smear, or nothing. Say if the programme recently moved to a different cycler. The molecular biology catalogue groups polymerase families, including those aimed at longer products. Use the quote request for the formulation. A minute-per-kilobase habit does not identify which bottle you need. If the length is one of several in the same tube, name the longest in the multiplex PCR enquiry reference. If the template itself is hard to denature or comes from a difficult extract, mention the nucleic acid analysis pathway so a short extension is not blamed for a template that never melted.
Questions from the bench
How long should a Taq-like extension be?
A widely used planning figure is about one minute per kilobase of amplicon, near 72 Celsius. It is a start for classical Taq-like enzymes, not a law for every polymerase. Engineered enzymes publish different speeds. Use the number on the card you opened, then confirm the full-length band on a gel.
Will a five-minute final extension fix a cycle that was too short?
No. The final hold can finish strands that were almost complete in the last cycle, and for Taq-like enzymes it can help complete an A overhang. It does not go back through earlier cycles. If every cycle extension was far too short, most molecules never included both primer sites and never became full templates.
What do partial products look like?
Often as a weak or missing band at the expected size, sometimes with a smear of shorter DNA. Those shorter strands are incomplete copies, not a second locus, when they shrink or vanish once the extension hold matches the length. A no-template lane should stay empty. A smear that is also in the no-template lane is a primer artefact, not a timing problem.
Do I time the extension from the insert alone?
Time it from the amplicon the primers actually copy, which includes both primer sequences and any vector sequence between flanking primers. A 2-kilobase insert amplified with primers that sit 200 base pairs outside it is not a 2-kilobase extension problem. In a multiplex, time the longest product.
References
Manufacturer names identify published method classes. Trademarks remain with their owners. Catalogue records on this site are independent references for enquiry. They are not a statement of inventory, distribution rights or a supply commitment. This page is educational. It is not medical advice, a diagnostic protocol or a biosafety approval.
Catalogue
Related products and categories
These links follow the subject of the article into published manufacturer references. A listing is a reference for an enquiry, not a statement of stock or distribution rights.
Continue in this cluster
Related reading
How polymerase chain reaction worksHow PCR copies a chosen DNA segment: the temperature cycle, reagents, controls and what a band on a gel does and does not prove.
A glossary of PCR termsPlan a PCR by defining amplicon, Tm, Ta, hot-start, plateau, master mix, NTC, inhibition, fidelity and multiplex before you mix.
Choosing an annealing temperatureWhy annealing temperature is not the Tm printed on a tube, and how salt, magnesium, additives, and a short gradient locate the real window.