I'm very familiar with how symbols display differently in inline mode (with $ ... $) vs. how display math shows them (with \[ ... \] or $$ ... $$). Two examples would be with limits

$\lim_{n\rightarrow \infty}f(x)$


\[ \lim_{n\rightarrow \infty}f(x) \]

and with sums

$\sum_{n=1}^{x} n^2$


\[ \sum_{n=1}^{x} n^2 \]

My Question is:

How do you display these things in inline mode (with all the nice inline formatting) as if they were in display mode ?

The inverse -- how to display things in display mode as if they were in inline mode -- is also of interest.


4 Answers 4


There are two separate aspects to your question, which can be addressed separately:

  1. How to control the size of the integral, sum and product symbols

  2. How to control the placement of the limits of integration, summation, and multiplication, namely side-set or above/below.

Default settings

The default settings, with the amsmath package loaded, are

  • Inline math

    • Small symbol size;
    • Limits are side-set for all operators;
    • The sumlimits and intlimits options of amsmath do not affect the placement of limits in inline math.
  • Display math

    • Large symbol size. See Tables 72 through 83 in the Comprehensive LaTeX Symbol List for the names of these 'large' or, more precisely, Variable-sized Math Operators;

    • For \int-type symbols, the limits are side-set unless amsmath was loaded with the option intlimits.

      The integral symbols are treated separately presumably because they are generally taller than the other variable-sized symbols.

    • For \sum, \prod, \coprod, etc the limits are set above and below the operator, except when the amsmath was loaded with the option nosumlimits.

      The sumlimits and nosumlimits options and the \limits and \nolimits commands affect the appearance not only of sum-type symbols but of \prod,\coprod, \bigcup and \bigcap, etc. as well.

Custom settings

  • To control the size of the symbol, one writes before the command generating the symbol

    • \textstyle for small symbols;
    • \displaystyle for large symbols; .
    • the declarations \textstyle and \displaystyle may also affect the behavior of subsequent commands in the current math-mode environment, as observed by @HaraldHancheOlsen.
  • To control the placement of the limits, one writes after the command generating the symbol

    • \nolimits for side-set limits;
    • \limits for limits set above and below.

These possibilities are illustrated in the table below:

enter image description here

The table was produced by the following code:

        >{$\textstyle}l<{$}     % first math column: text style
        >{$\displaystyle}l<{$}} % second math column: display style
Placement of limits & \multicolumn{2}{c}{Size of operators} \\
 & \multicolumn{1}{l}{small:}
 & \multicolumn{1}{l}{large:}\\
 & \multicolumn{1}{l}{\texttt{\textbackslash textstyle}}
 & \multicolumn{1}{l}{\texttt{\textbackslash displaystyle}}\\
Next to symbol: & \multicolumn{1}{l}{\phantom{\texttt{\textbackslash displaystyle }}}\\ 
\texttt{\textbackslash nolimits} 
         &   \sum\nolimits_{i=1}^N a_i 
         &   \sum\nolimits_{i=1}^N a_i \\[2ex]
         &  \prod\nolimits_{j=0}^J k_j
         &  \prod\nolimits_{j=0}^J k_j \\[2.5ex]
         &   \int\nolimits_{-\infty}^\infty f(x)\,\mathrm{d}x
         &   \int\nolimits_{-\infty}^\infty f(x)\,\mathrm{d}x \\[5ex]
Below \& above symbol:\\[-1ex]
\texttt{\textbackslash limits}
         &   \sum\limits_{i=1}^N a_i 
         &   \sum\limits_{i=1}^N a_i \\[3.5ex]
         &  \prod\limits_{j=0}^J k_j
         &  \prod\limits_{j=0}^J k_j \\[4ex]
         &   \int\limits_{-\infty}^\infty f(x)\,\mathrm{d}x
         &   \int\limits_{-\infty}^\infty f(x)\,\mathrm{d}x \\

Personal comments

Finally, some personal views on the (ab)uses of the \limits and \displaystyle commands when in inline math mode:

  • It's generally not a good idea to use the \limits command when in inline math mode. Otherwise, one is virtually assured of wrecking the appearance of the paragraph where the formula appears.
  • When in inline math mode, it's frequently not even necessary to indicate the full set of limits of a summation or multiplication. Expressions such as \sum_i or \prod_j are usually just fine. You may even be able to get away with omitting the i and j indices of summation/multiplication.
  • Using the \displaystyle command (to force the creation of large symbols) while in inline math mode is an even worse idea than using \limits.
  • 16
    Informative & educative!
    – Werner
    Commented Oct 26, 2011 at 23:07
  • 7
    Minor nit: You might have made it more clear that \displaystyle and \textstyle do not affect only the following symbol. They are declarations whose effects last until the end of the formula (or group) or until overridden. Commented Oct 27, 2011 at 16:20
  • 16
    I have found that \displaystyle in inline math is useful when preparing beamer presentations. In an itemized list, sometimes there is a short text (like two words) preceding a formula, and I don't want to waste valuable vertical space on a displayed formula but do want large symbols. Then \displaystyle is appropriate, since the items in a list are visually similar to displays.
    – Ryan Reich
    Commented Feb 14, 2012 at 19:47
  • 3
    @RyanReich: Good point about Beamer presentations' material often having different requirements than that of "ordinary" inline material. Thus, bullet point stuff can behave very much like display math material -- and hence may need to invoke \displaystyle and/or \limits.
    – Mico
    Commented Feb 14, 2012 at 23:12
  • @XavierStuvw - Many thanks for the edits you applied. After almost nine years, it was definitely time for a visual refresh!
    – Mico
    Commented Aug 5, 2020 at 17:44

The other answers are excellent, and normally I would not try to change the behavior of inline functions. However, there is one case where I wish inline functions to behave just like displayed functions. That is when I write exams. I am willing to compromise on typesetting for readability. If you put


in your preamble, every equation will be typeset in that manner, and you don't have to put \displaystyle in every equation. Of course, if you are looking to change just a few equations, then \displaystyle is easy to use.

  • 5
    the only answer that actually solves this problem
    – Fraïssé
    Commented Aug 20, 2015 at 4:23
  • 5
    @IllegalImmigrant Unfortunately, it's the wrong answer. It has several upvotes, probably because of simplicity; but this simplicity is at the expense of good typography. No, doing \everymath{\displaystyle} is generally a bad idea.
    – egreg
    Commented Oct 15, 2015 at 12:20
  • 1
    Excellent! When writing worksheets and exams, this is exactly what I need!
    – Andrea
    Commented Jun 18, 2022 at 20:03

Good typography relies on the balance of all aspects of black and white on the page. Uniformly spaced lines for the ordinary text make for good legibility. That's why some symbols, that in display are rendered with limits above and below, are set with limits on the side when used in in-line formulas.

I too, when I began to use TeX, tried to set limits for sums above and below also in in-line formulas, but I soon realized that it's wrong: two white bands separated that line from the next ones.

For that very reason the symbols for summation and integral are set in two different sizes: a smaller one for in-line formulas, a larger one for displayed formulas that don't have spacing constraints.

Displayed formulas are set using (automatically) \displaystyle, so

$\displaystyle\sum_{n=1}^{x} n^2$

would produce the same symbol used in displayed formulas and set limits above and below. But this will damage the balance of the page beyond repair. A less invasive construction

$\sum\limits_{n=1}^{x} n^2$

and its sibling


will do less damage to the page, but will nevertheless spoil it.

Such constructions have their use, for examples in tables where TeX would use in-line math mode. But I will never suggest to use them in normal text.

  • 4
    While true, your answer reads a little patronizing. I think we can treat users of this site as adults. if somebody asks how to do something, we should not assume that they do not know what they are talking about, or at-least should not let our answer show that... :-) Commented Oct 26, 2011 at 21:32
  • 1
    I agree... if you would re-word it to be less condescending (and actually answer the question instead of saying how wrong it is to ask it) I would likely up-vote your answer.
    – akdom
    Commented Oct 26, 2011 at 21:39
  • 26
    @YossiFarjoun, while egreg's response is a bit sharp, I don't think it is always a bad idea to question someone's purpose for wanting to achieve a certain look with TeX. As someone who started using LaTeX not too long ago, I must credit my still-improving sense for good typesetting to people who told me that I didn't actually want to do what I said I wanted to do. I would be a poorer TeXer were it not for them. In general, it seems to be a feature of well-designed systems that learning them requires shifting one's approach to problems. Commented Oct 26, 2011 at 22:08
  • 2
    That's a fantastic new version egreg. Definitely a more complete answer to the question.
    – akdom
    Commented Oct 27, 2011 at 23:37
  • 1
    @dh87 That's simply wrong, particularly in exponents.
    – egreg
    Commented May 1, 2016 at 15:36

The solution you seek is to use the \displaystyle command within the inline environment as such.

$\displaystyle\sum_{n=1}^{x} n^2$

This will give the nice effect of the starting term being underneath the sigma and the maximum value above while keeping the symbols inline.

  • 14
    Interesting how the presentation is done in the 3rd person.
    – Werner
    Commented Oct 26, 2011 at 20:55
  • 1
    I figured that it would be the most immediately accessible for people quickly looking for an answer :) .
    – akdom
    Commented Oct 26, 2011 at 20:58
  • 1
    Well, it a wrong answer.
    – egreg
    Commented Oct 26, 2011 at 21:00
  • 1
    @egreg: The answer isn't wrong....it answers the question exactly. I think that the question is wrong...but that's what down-voting the question is for... Commented Oct 26, 2011 at 21:33
  • 19
    I have to admit that I don't like the way that the answer is phrased (irrespective of whether or not it is correct). Answering ones own question is absolutely fine, but do it honestly not pretending that you aren't the person who asked the question. Commented Oct 27, 2011 at 6:36

You must log in to answer this question.

Not the answer you're looking for? Browse other questions tagged .