I am learning HTML and CSS, discovering that browsers implemented rounded corners and shadows using CSS... just about the time that UI went to "flat / sharp corners" without shadows, etc. It seems sad that right when this long-sought ability arrived, we had lost interest.

I came to ask this after seeing questions and answers on this site which linked to articles about rounded corners being "easier on the eyes" (linked in an answer to this question, reminiscent of the decades-long raving about serif and sans-serif) and about "women preferring rounded corners" (and they're more than half the human race and control 63% of spending decisions so we don't want to alienate them)...

Here is a reference for what I am saying: Nielsen Norman Group article

The classical dictum implies that beauty in design results from functionality, and thus, aesthetic considerations in design should be secondary to functional considerations. Designers should focus on elements that are critical to functionality, and only after those have been identified can they start searching for the most beautiful implementation that accommodates the functionality constraints.

The point is, is this field based on research and engineering, or does it blow in the wind? If rounded corners are better, then by God we should stand up for them. Else we should realize and accept that our conclusions of 'betterness' are simply rationalizations after the fact, like most of what people do. That is not wrong, but stop calling UX a science. (and most everything else too)

I didn't write the research, or the other questions and answers on the site, I am just asking if I should take these things seriously?

  • 1
    It's an interesting question but there may not be an objective answer.
    – Mayo
    Commented Nov 29, 2016 at 21:54
  • 1
    Personal opinion, yes there are trends, but nothing stops you using rounded corners, gradients, drop shadows if it pleases the stakeholders, they insist on it or it is part of the design guide for the business. Commented Nov 29, 2016 at 22:01
  • 1
    "if they are better" = citation? It's really just an aesthetic call. Aesthetic trends come and go. CSS features usually follow the prevailing visual trends so are always a bit late to the party.
    – DA01
    Commented Nov 29, 2016 at 22:04
  • 1
    Straight edges usually render better across desktop, TV, and mobile, especially on lower-res displays. Plus excessive rounded edges may come off as Apple-like.
    – Alan
    Commented Nov 29, 2016 at 22:13
  • Who is this "we" that you say proclaimed rounded corners must be used for the sake of UX? I think a lot of folks would consider button shapes to be more of a design issue. Can you back up your assertion that the prevalence of rounded corners is even considered to be a "UX decision"?
    – Tim Grant
    Commented Nov 30, 2016 at 21:20

6 Answers 6


Rounded Corners are best for bestest buttons!

Take, for example, the primary button in And®oid, the Floating Action Button™.

A close examination of the button reveals it to be a rounded square:

enter image description here

There are no circles in modern design, only fully rounded squares.

  • i.sstatic.net/ve1RL.jpg?s=32&g=1
    – Devin
    Commented Nov 30, 2016 at 23:19
  • Thus, the thing called Flat Design.
    – user67695
    Commented Dec 1, 2016 at 14:33
  • 1
    Knowing how and why "Flat Design" became a thing goes all the way back to understanding how die cutting, Flash and printing, amongst many other technologies, commercial imperatives and technological constraints on mediums of production, presentation, publication and packaging influence outputs. Look at how we are/were limited by respective technologies of multiple mediums in all their facets of creation and exploration. Limitations and commercial interests play much more important roles in shaping trends than pure functionality desires of consumers and the urges of creatives. @nocomprende
    – Confused
    Commented Dec 1, 2016 at 14:57

If this question (whether or not our conclusions are subjective) is opinion-based, then I think that shows that the Field of UX is opinion-based.

Let’s start here. UX is a huge field of study. Like a lot of fields, there are objective and subjective aspects to it. In terms of a GUI, debating rounded v. square corners is mostly a subjective visual design decision. One could objectively test customer preferences and such, but at the end of the day, some art director is going to make a call on it.

Whether the visual UI has rounded corners or not may very well play a larger role in how usable the UI is overall. Then again, it may have zero impact. Again, this could be tested objectively, but is usually going to be low priority in terms of the total UX objectives.

Another thing to mention is that while groups like the NNG are highly skilled and have a lot of great data and opinions, context is everything. They are good to listen to, to get an understanding of the foundational rules and guidelines, but one should never adhere to what they say as literal gospel as context is everything and one simply can’t create UX roles that apply to all situations equally.

The concerns you have about articles that say something like “women preferring rounded corners” is a very valid one. We work in a field that tends to (this is my opinion here:) latch on to trivial research to make decisions that really need to take into account many more factors than one particular study of one particular UI.

That said, that doesn’t mean those studies are useless…it’s just that we need to treat them as what they are…little bits of data that can help contribute to a solution—it’s just that they likely shouldn’t dictate a solution.

  • Thank you. The only problem I have is that due to the prevalence of rounded corners in (now woefully out-of-fashion) websites, the CSS standard was actually revised and browsers modified, to cater to that. As an engineer, I would only go to that much effort if rounded corners were here to stay, and no way, no how would they turn out to just be a preference (unless it was one that every single person shared). Browsers are infrastructure, and standards are like laws: they should not change willy-nilly. We don't make the street signs pink and purple no matter how many people want that.
    – user67695
    Commented Nov 30, 2016 at 20:09
  • @nocomprende well, they don't change willy nilly. The W3C is a behemoth of a committee. Many web developers would say standards have changed slowly and thoughtfully. But remember, we're not talking visual design standards. We're merely talking implementation standards. The browsers don't dictate the former. But they do dictate the latter.
    – DA01
    Commented Nov 30, 2016 at 20:49
  • But why bother to alter the standards for something that will just be dropped later?" Let trends and fashion burn out, don't enshrine them in law, "to hang quite out of fashion, like a rusty mail in monumental mockery." Why did we get played for the fool by writing up articles and debating the merits of something with no more standing than a spread in Vogue magazine?! Ignore these things.
    – user67695
    Commented Dec 1, 2016 at 14:27
  • @nocomorende the world
    – DA01
    Commented Dec 1, 2016 at 16:48

I really don't understand (no comprende!) where do you get this from. If anything, sites are using rounded shapes more and more. Bootstrap and Material are between the most used frameworks/guidelines, and most elements are rounded.

Also, not sure what does it have to do with CSS. CSS will render whatever you instruct it to do. If you want sharp corners, border-radius:0 will suffice, or border-radius:(n) for anything rounded

However, there are certain elements that tend to be sharp (usually containers). This is because of structural perception, people tend to think straight lines are "safer" and simpler. Thus, it's common to see a straight lined container containing a rounded-corner button .

As for UX not being a science... well, you're correct. Otherwise, this would be the first time I hear that UX is a science. In any case, it's a discipline that takes from many different sciences. And those sciences include scientific methods to derive ever-changing user preferences. Anything that includes the word "user" will be subjective by default, but you can measure that subjectivity and get information from an statistical average (that will NEVER cover 100% of cases!)

From Experience Design

The field of user experience design is a conceptual design discipline and has its roots in human factors and ergonomics, a field that, since the late 1940s, has focused on the interaction between human users, machines, and the contextual environments to design systems that address the user's experience. With the proliferation of workplace computers in the early 1990s, user experience became an important concern for designers. It was Donald Norman, a user experience architect, who coined the term "user experience," and brought it to a wider audience.

  • My point is that experience should result in a more common, shared way of doing things, like how cars all have steering wheels (the joystick experiment didn't work) and they all have controls mounted on stalks, and doors that hinge from the forward edge... The other tries didn't work. So, if someone thought that rounded corners were important enough to build in to the browser and standards, it should not just vanish later in a marketing pivot, to flat. If it is really important enough to modify the infrastructure, it should be that, like a steering wheel, it is just the only sensible option.
    – user67695
    Commented Dec 1, 2016 at 14:25
  • 1
    I'd suggest a more direct, handle bar-like, mapping of steering to action would have been, and remains, a better choice than ship-like steering wheels, from a UX point of view, for turning a tire vehicle. But the technical limitations of high gearing requirements to cover the lock-to-lock angle of the wheels, the lack of powered steering for decades, and the then subsequent ubiquity of the ship-steering paradigm - across generations - meant we're not returning to handle bar-like direct control in anything other than racing cars where the immediacy is imperative to performance and control boost
    – Confused
    Commented Dec 1, 2016 at 16:23
  • @Confused nice exploration of my analogy... Perhaps part of the difference is that race cars go really fast, and so you need precise control of very small amounts of turn, whereas passenger cars might need to make extreme turns at very low speeds, like parking. So the wheel is more general-purpose. It is so much fun to loosen my grip and let the wheel spin through my hands as it returns to center after taking a corner. That alone makes it worthwhile. I wish to God that they would put the high-beam switch back under my left toe where it belongs!
    – user67695
    Commented Dec 1, 2016 at 17:18

Please read the top rated answer on this question (related) - How do rounded corners affect usability? and read the book "Humans prefer curved visual objects"

We are preconditioned over time to what we find in the natural world. Curves are more likely found in nature, while square corners are pretty unlikely, and often where they might occur (eg due to fractures, breaks or other forces) they get worn down over time to make a natural curve again. Square corners simply do not have a place in natural world.

  • This is actually a good example of what @nocomprende is getting at...we put way too much weight into little bits of research like this.
    – DA01
    Commented Nov 30, 2016 at 23:41
  • Because humans prefer them, all the UIs still use curved objects.
    – user67695
    Commented Dec 1, 2016 at 14:20

Every shape should be the way it is to serve a purpose, or for a reason. "Form follows function".

"Rounded corners are better". This is the main point of your question. No predefined shape is better than any other. A shape/form is good as long as it solves the function of the object. And it is "better" the more effectively it solves this function.

This doesn't imply that there is an archetype form for each object, as alterations in a shape might serve additional functionalities.

Take the case of a shape which solves the object's functionality:

  • If alterations to the original shape are made purely for aesthetic reasons (without serving any additional functionality) it is not better than the original shape.

  • If alterations to the original are made to serve additional functionality: it might be better as it solves more actions; or it might be cluttering the main function.

For physical objects sharp corners might be dangerous, we could agree some roundness in corners might be safer. How much roundness is perfect?

enter image description here

There is no canonical answer but maybe we could agree:

  1. The first shape serves correctly a smartphone's reading functionality.
  2. The second shape serves correctly an smartphone's reading functionality. Adds an extra functionality which is safety in the outer corners.
  3. The third shape is similar to the former but might start to disturbe the main functionality.
  4. The fourth shape looks like it is not serving correctly the main functionality.

Digital objects resemble physical objects. Our brain is used to understand that sharp outer corners can be dangerous, so some roundness looks safer to the eye.

  • Thus, we love the dangerous look of Flat Design.
    – user67695
    Commented Dec 1, 2016 at 14:34
  • Flat design is more about the fills and shadows. Even on material design you can find rounded corners.
    – Alvaro
    Commented Dec 1, 2016 at 14:42

Material design suggests using very subtle contours as opposed to large rounded, possibly clumsy looking, corners. Surely it moves with the times though, you dont see many bubble cars driving around these days!

Why Rounded Corners Are Easier on the Eyes

  • Hope you don't mind me adding the text from the referenced page. If "there is more to it than that" (and the reference explains the physiological reasons why) then we should stick to our guns. Better is better, right? If people said "we just don't like seatbelts" we would override them. We do. It's not just a good idea, it's the law (of perceptual quality).
    – user67695
    Commented Nov 30, 2016 at 14:24
  • A link in your reference goes to: basement.org/2005/11/why_do_we_love_rounded_corners.html where it says, "I would venture that our attraction to rounded corners goes beyond the aesthetic and speaks to something more." And there are lots of comments. One commenter even says, "Are rounded corners ever going to go out of fashion?" Hmm.
    – user67695
    Commented Nov 30, 2016 at 14:31

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.