When specifying a 0 value in CSS, should I explicitly mark the units or omit?

Solution 1:

I argue you should also omit the units.

From a programmer's perspective, 0 == null == none == false, where 0px == 0px only.

Which means that if you specify a border width of 0 then no border will be there, but if you specify a 0px border, then a border of 0 px will be created (that's the idea behind it, in reality 0px gives the exact same result like 0).

Further Points

unit-less 0 makes it easier to read as it is easily distinguishable from normal unit'ed values.
It makes sense to remove the units as they have no point in being there (0 could mean size, color, etc.).

Conclusion: Omit the units in 0. They're not needed and confusing.

Solution 2:

As a note to this question: Unitless 0 as length won't work in calc() and other math functions. You have to write 0px. And things would be more buggy if the unit omitted 0 is taken from some css variable.

Specification:

Note: Because <number-token>s are always interpreted as <number>s or <integer>s, "unitless 0" <length>s aren’t supported in math functions. That is, width: calc(0 + 5px); is invalid, because it’s trying to add a <number> to a <length>, even though both width: 0; and width: 5px; are valid.

.a, .b { border: 10px solid; height: 30px; }
.a { width: calc(300px + 0px); border-color: #f00; }
.b { width: calc(300px + 0); border-color: #330; }

<div class="a">width: calc(300px + 0px);</div>
<div class="b">width: calc(300px + 0);</div>

I would suggest always use 0px when you are writing CSS variables. So it won't make you and others confuse when they are trying to use the variable in some calc() functions.

Solution 3:

The CSS specification says (quote):

After a zero length, the unit identifier is optional.

And I've read many times that it is suggested to omit the unit, I can't remember where.

So omit them.

Solution 4:

I also like to put units by my zeros because ...

IDEs like to have the values there so they can increment them properly, as Istvan mentioned.
I don't have to type the unit later if I put another value in. And I don't have to try and remember to remove the unit if it's zero or put it back if it's not.
One poster said the naked zero was more readable. I find that it's not. When there is a unit, you have context, and that makes it more readable. For example, if I come across a percent sign, I know I use that in different ways than the "px" unit, so I instantly know some of the things it could be.
Another guy, "red", had this to say, which looks like good info. Here's a snippet...

However, you will notice it if you intended to change a prior value in the cascade. a '0em' says, 'change the prior value to '0em', but a zero may simply say, 'disregard this rule' and leave the prior rule in effect. This may not be at all what you intended with your naked '0'. original article link