Detecting viruses: the plaque assay

One of the most important procedures in virology is measuring the virus titer – the concentration of viruses in a sample. A widely used approach for determining the quantity of infectious virus is the plaque assay. This technique was first developed to calculate the titers of bacteriophage stocks. Renato Dulbecco modified this procedure in 1952 for use in animal virology, and it has since been used for reliable determination of the titers of many different viruses.

polio-plaquesTo perform a plaque assay, 10-fold dilutions of a virus stock are prepared, and 0.1 ml aliquots are inoculated onto susceptible cell monolayers. After an incubation period, to allow virus to attach to cells, the monolayers are covered with a nutrient medium containing a substance, usually agar, that causes the formation of a gel. When the plates are incubated, the original infected cells release viral progeny. The spread of the new viruses is restricted to neighboring cells by the gel. Consequently, each infectious particle produces a circular zone of infected cells called a plaque. Eventually the plaque becomes large enough to be visible to the naked eye. Dyes that stain living cells are often used to enhance the contrast between the living cells and the plaques. Only viruses that cause visible damage of cells can be assayed in this way. An example of plaques formed by poliovirus on a monolayer of HeLa cells is shown at left. In this image, the cells have been stained with crystal violet, and the plaques are readily visible where the cells have been destroyed by viral infection.

The titer of a virus stock can be calculated in plaque-forming units (PFU) per milliliter. To determine the virus titer, the plaques are counted. To minimize error, only plates containing between 10 and 100 plaques are counted, depending on the size of the cell culture plate that is used. Statistical principles dictate that when 100 plaques are counted, the sample titer will vary by plus or minus 10%. Each dilution is plated in duplicate to enhance accuracy. In the example shown below, there are 17 plaques on the plate made from the 10-6 dilution. The titer of the virus stock is therefore 1.7 x 108 PFU/ml.


Next we’ll consider how the plaque assay can be used to prepare clonal virus stocks, a step that is essential for studying viral genetics.

Dulbecco, R., & Vogt, M. (1953). Some problems of animal virology as studied by the plaque technique. Cold Spring Harbor Symp. Quant. Biol., 18, 273-279

Comments on this entry are closed.

  • Raj 7 July 2009, 6:49 am

    Thank you for the post. Plaque assays can be done to determine infectious virus titer for those viruses whose susceptible cells grow in monolayers however some viruses are grown in suspension because of the nature of the host cell e.g. HIV in H9 cells, HHV-8 from induced KS-1 cells. For these viruses, infectious virus titer cannot is determined by plaque assay rather by TCID50. Now, TCID50 is not absolute quantitation of infectious virus unlike plaque assay which is absolute. TCID50 titer is calculated by monitoring viral CPE which is not always easy because some viruses do not produce CPE. In these cases, how can one calculate infectious virus titer?
    Thank you.

  • Gabriel Marceau 9 July 2009, 1:10 pm

    TCID 50 is not an assay per se but a calculation method. It can be used if CPE is present but it's usefulness is more evident when no CPE is present, thus the need for an alternative method to plaque assays (a plaque is a CPE). In the case of human coronaviruses (HCoV) an immunoperoxydase assay can be done in 96-well plaque with dilutions of the virus and the TCID calculated based on wich well contains infected (positive) cells ( No CPE is detected, only the presence of viral antigens on the surface of the infected cells.

    A further note, the mouse coronaviruses (MHV) can be detected by plaque assay despite the fact that they are not “lysogenic”. They wont kill the infected cells per se but induce the fusion of neighbouring cells. The scincytia they form can either be seen using a microscope or by the plaque assay (the scincytia are fragaile and when pouring the fixating agent they get dislodgedm, leaving a plaque).

  • btomas 12 July 2009, 12:06 pm

    Raj. About your question, for quantifying infectious viruses you can use immunofluorescence assays and count the focus of infection or currently there are new works to improve the values obtained by qPCR to determinate the infectious viruses, for example with a pre-treatment to eliminate viruses with structural damages, quantifying only infectious viruses.

  • marlakins 10 August 2009, 3:15 pm

    Vince, do you think that quantitative PCR is an accurate means of comparing influenza virus between different samples, say in a controlled experiment? It seems that most papers stick with plaque assays.

  • profvrr 10 August 2009, 5:39 pm

    Q-PCR measures viral RNA, not infectivity – a plaque assay is the best
    measure for infectivity. So it depends on what you want to measure.
    Personally I find plaque assays far easier to carry out than Q-PCR.

  • Shorts 21 November 2009, 9:46 pm

    Hi, I just want to know the formula used for the determination of the PFU? Additionally, isn't the stock concentration relevant to the dilution? What stock concentration are you assuming was present there?

  • profvrr 22 November 2009, 7:04 am

    The formula for calculating plaque titer is: (number of plaques)
    divided by (dilution of the stock virus on the plate that was
    counted). In the example in the figure: (17 plaques) divided by
    (10e-8) is 1.7 x 10e8 pfu/ml. If you plate a wide range of dilutions
    you don't need to know the concentration of the stock – you only need
    to know how much to dilute which is related to the maximum virus titer
    expected. In this example, the virus is known not to exceed 10e9
    pfu/ml so the highest dilution need only be 10e-8.

  • beehuiyeo 11 January 2010, 10:14 pm

    Actually i would to detect the cyanophages in Singapore. I found that, I cant get plaque but I get clear or lawn. Is it also because of cyanophage present. Thanks.

  • profvrr 12 January 2010, 4:57 am

    If the lawn is cleared then you should be able to get plaques. Have
    you tried many dilutions? Different agar?

  • beehuiyeo 13 January 2010, 7:27 pm

    Thanks for your comment. Is is the whole agar (clear) as a plaque? I have not tried on many dilutions, I will try on it. But, why the different agar can give different result? I was trying on the filamentous cyanobacteria phages, do you think this type of bacteria phages able to give small plaque?? Thanks.

  • profvrr 14 January 2010, 4:47 am

    DIfferent types of agar (oxoid, noble, bacto, agarose) have different
    components that may inhibit plaque formation. For example, certain
    rhinoviruses will form plaques only in overlays containing type VII
    agarose, not any other type of agar. Yes, there is no reason why a
    filamentous phage cannot form a plaque. If the phage clears a lawn of
    bacteria then on higher dilution you should see plaques.

  • Hong-Wai Tham 26 January 2010, 6:22 pm

    Great explanation !! I like it.

  • ravidhakar 31 January 2010, 3:02 am

    DIfferent types of agar (oxoid, noble, bacto, agarose) have different
    components that may inhibit plaque formation. For example, certain
    rhinoviruses will form plaques only in overlays containing type VII
    agarose, not any other type of agar. Yes, there is no reason why a
    filamentous phage cannot form a plaque. If the phage clears a lawn of
    bacteria then on higher dilution you should see plaques

  • Cenef 16 February 2010, 7:24 am

    I'm trying to get information on the optimum volume (or thickness) of the infecting aliquot in the plaque assay. In the above example you use 0.1 ml. As the thickness of this liquid layer affects the probability of virus particle attaching to cells during the adsorption period, it should be matched with the surface area of the monolayer, adsorption time, and temperature. What literature is out there on these factors, and is this virus specific?

  • Claudia 19 February 2010, 10:38 am

    Get help!! Thanks

  • Claudia 19 February 2010, 10:39 am

    Get help!!! Thanks

  • murine 24 March 2010, 11:05 pm

    Hi Dr Vince,
    In case of emerging Picornaviruses, the virus with less adaptation to cell culture (which improves upon 2-3 passages), which method could be more suitable for quantifying the infectious viral particles?
    And Thank you for the Virology Infection. Which appears to be like that of HIV.

  • profvrr 25 March 2010, 4:54 am

    If the virus causes cytopathic effect in cells it can be titrated
    using an endpoint assay, such as 50% tissue culture infectious doses
    (TCID50). This is the amount of virus needed to cause cytopathic
    effect in 50% of infected cultures. See….

  • megha 5 April 2010, 3:24 am

    how to stain the plaques, I mean do we need to add agar containing stain in nutrient medium to the plate?

  • megha 5 April 2010, 10:24 am

    how to stain the plaques, I mean do we need to add agar containing stain in nutrient medium to the plate?

  • shirod 6 May 2010, 2:25 am

    it seems like the procedure is too complicated i am a new micro student , is there a simpler way to do plaque assay? or ways to improve the procedure to get a more readable and countable result?

  • profvrr 6 May 2010, 2:40 pm

    Depending on the goal, the stain can be either added to the overlay,
    or added to the cells after the overlay has been removed. For example,
    neutral red dye can be added to the agar overlay to visualize plaques
    more easily for isolation. If the goal is not to isolate plaques, but
    merely to count them, then the monolayer can be stained after the agar
    is removed with a dye such as crystal violet.

  • profvrr 6 May 2010, 2:41 pm

    The plaque assay is not complicated; it is one of the simplest, and
    most informative procedures in virology. I'm making a video of the
    entire process which will show how easy it is.

  • Lyndsinz_89 18 June 2010, 8:03 pm

    Hi Sir, good day! im a 4th year BS Chemistry student. For us to graduate, we are required to have an undergraduate thesis. The topic that I want to study is entitled “The Antiwart Activity of Ficus septica”. My problem is, how can I culture this kind of virus which is human papillomavirus (HPV) in a simplest yet best method? What is the best assay that I can do to detect the presence of this virus? I will deeply appreciate your help Sir? Thank you very much. God bless and more power.

  • T4phage 1 July 2010, 1:21 pm

    I'm currently trying to enumerate very low levels of T4 bacteriophage using the plaque assay and I've found that the limit of detection may not be high enough. WHat is the limit of detection of the plaque assay and is there another assay method I could use with a lower limit of detection?


  • profvrr 1 July 2010, 6:34 pm

    The detection limit is roughly 10 PFU per ml, if you are plating out
    0.1 ml per plate. You could use a limit dilution method (such as 50%
    infectious doses) which involves more assay plates and can therefore
    detect fewer than 10 PFU per ml, but the readout is cell killing, not
    plaquing. Have you considered nucleic acid detection methods such as

  • T4phage 5 July 2010, 10:43 am

    Thank you, my next step was to look at other detection methods such as PCR but I was unsure of the detection limits for this assay also, should PCR be able to detect even 1 viable phage? Thanks

  • profvrr 8 July 2010, 2:23 pm

    PCR will not give information on viable phages, only on the presence
    of phage DNA. It should be possible to refine the PCR to detect one

  • Juan Carlos 29 July 2010, 4:23 pm

    Hi Dr Vincent,

    I’m actually having problems with the variability of the plaque assay. I have repeated the assay several times to determine the concentration of my viral stock and in occasions the variability is two-fold and even greater.
    Is there an inherent variability of the technique or is it me to blame?
    Thank you in advance (great informative website by the way).

  • profvrr 29 July 2010, 5:13 pm

    In a plaque assay, a two-fold variability is routine due to errors in
    pipetting, dilution, etc. Five-fold would not be acceptable.

  • dracolll 3 September 2010, 11:00 pm

    Hi Sir, I'd like to know is there any reason, other than preventing contamination, that when we perform plaque assay to titer influenza viruses, we need to invert the plates? Some say it's due to the polarity of influenza virus growth. But that's not quite convincing.

  • Cassie 13 October 2010, 10:57 am

    Dear Dr Vincent,

    I'm currently doing plaque assay using Vero cells to quantify my Chikungunya virus. I'm now optimizing the incubation period prior to staining. I can see plaque formation 2 days PI via naked eyes. My question is, is it too soon for me to stain the plaques? Or should I incubate longer? Because when I tried incubating them for 3 days but the plaques formed are large and start to overlap each other. I appreciate your help. Thanks in advance.

  • profvrr 14 October 2010, 3:00 pm

    If the plaques overlap after 3 days incubation then you should

    certainly stain after two days.

  • Saad Saleem 19 October 2010, 4:24 pm

    Hi Dr. Vincent,

    I am looking for ways of optimizing plaque assay. In terms of time, or the quality of results. Do you have any suggestions

  • Tlawson 30 October 2010, 8:36 pm

    Hello. I am a biochemist on sabbatical leave in a picornavirus laboratory learning some basic virology techniques. I am working with EMCV but am having trouble with the plaque assay procedure. The lab uses agarose overlays, removes the overlays after incubation, and then fixes/stains with crystal violet/MeOH/H2O solution. This seems to work fine with HeLa cells. When I use this procedure with EMCV-infected L929 cells, however, the cells come off the plate with the overlay. I can see some clear plaques in the wells, and I can see the intact monolayers of cells under a ‘scope before I remove the agarose plugs. Are there any suggestions for keeping the cells attached to the well surface instead of peeling off with the overlay?

  • profvrr 1 November 2010, 3:32 pm

    Before peeling off the agarose overlay, add enough 10%
    tricholoroacetic acid to cover the overlay and incubate 10 minutes.
    This will fix the cells to the plastic so they do not come off. We use
    this step routinely for all of our plaque assays.

  • Tlawson 1 November 2010, 9:21 pm

    Thank you! We ordered TCA today.

  • Shalini 23 November 2010, 8:19 am

    I am trying to do a plaque assay for JEV but I could not get any plaques.
    Can you please suggest me some reasons as to why this is happening??
    can JEV be inactivated due to some reasons??

  • profvrr 30 November 2010, 9:35 pm

    There are many factors that determine whether or not a virus will form
    plaques in a cell line. It’s important to know that the virus is
    cytopathic in the cells, but that alone is not sufficient. The easiest
    variable to change is the overlay: there are many different types
    (agar, agarose, methylcellulose) and they should be evaluated

  • D_diston 4 December 2010, 7:51 pm


    Do you have a reference for this detection limit?


  • Sujit pujhari 11 January 2011, 6:19 pm

    May I know in which cell line you are using to titrate JEV.

  • Drgvmd 11 April 2011, 9:38 am

    I am trying to do a plaque assay with chikungunya but this virus does not produce a clear cut CPE. Kindly let me know if I can proceed with the plaque assay

  • profvrr 13 April 2011, 6:50 pm

    I would proceed…sometimes CPE isn’t a predictor of good plaquing.

  • Vinniethepolar 18 April 2011, 4:37 pm

    I have the same problem of having cells come off the plate with the overlay. Besides TCA, any other chemical you can recommend? Thanks very much

  • profvrr 18 April 2011, 4:43 pm

    Try 10% methanol.

  • Curits 13 May 2011, 4:50 pm

    You will only be able to assess the antiwart activity on DNA replication by assaying for viral copy number. You cannot reasonably assess effect of any compound on production of virions because virion production of HPV has no cell model and is not cytolytic.