When using JavaScript to check a form, the first thing you need is the onsubmit event handler.
This event handler specifies a script that is executed when the user submits the form. That script is where
you check whether certain fields have a value, whether the user has checked at least one checkbox,
and any other checks you deem necessary.

The general syntax is:

<form action="something.pl" onsubmit="return checkscript()">

where checkscript() is the name of the script. This script should return either true
or false. If false is returned, the form will not be submitted. If either true
or false is returned the script stops.

So the general script becomes something like:

function checkscript() {
 if (some value is/is not something) {
  // something is wrong
  alert('alert user of problem');
  return false;
 }
 else if (another value is/is not something) {
  // something else is wrong
  alert('alert user of problem');
  return false;
 }

 // If the script makes it to here, everything is OK,
 // so you can submit the form

 return true;
}

Of course this function can become much more complex if you have to check a complicated
form with a lot of radio buttons and things. The general idea remains the same, however: You
go through the elements, check whatever you want to check and as soon as you find any mistake, you return false, after
which the script stops and the form is not submitted.

Once you've found a mistake, you should notify the user of the problem. This used to be done by an alert,
but nowadays you can generate error messages and insert them next to the form field.

Only at the very end of the script, when you have checked all elements and encountered no mistakes,
you return true, after which the form is submitted.

The form validation script needs to access the form in the HTML page to determine
what values the user has filled in. So first we have to enter the form by means of the
Level 0 DOM.
The general syntax for accessing a form element is:

document.forms[number].elements[number]

When the page is loaded, JavaScript makes an array forms in which it
puts all the forms that are on the page. The first form is forms[0], the
second is forms[1] etc.

Each form has another array in which JavaScript puts all the elements in the form.
The first elements is elements[0], the second elements[1] etc.
Every <input>, <select> and <textarea> is an element.

In some cases, it's better to use the names of the forms and elements. In HTML, you have
to give a name to each form and each element, like:

<form name="personal" action="something.pl" onsubmit="return checkscript()">
<input type=text size=20 name=name>
<input type=text size=20 name=address>
<input type=text size=20 name=city>
</form>

Now you can access these elements by:

document.personal.name
document.personal.address
document.personal.city

The advantage of using names is that you can put all elements somewhere else in the page
and still have a working script, while a script using numbers will have to be changed.
After all, the input box city is document.forms[0].elements[2] in
the example above, but when you suddenly put it at the top of the form, it becomes
document.forms[0].elements[0] and you have to change the script.

When you start writing your own scripts using the code snippets below, it is
always very important to know exactly how your form is built. To help you, I
wrote a form printing script that
prints out the form structure for you.

JavaScript has a few built-in methods and properties for dealing with forms. Three of them
are especially important:

You can submit a form by using the submit() method. To submit the first form on
the page, do

document.forms[0].submit()

Please note that when a form is submitted by JavaScript the onsubmit event handler
is never executed.

To reset the form, do

document.forms[0].reset()

I assume, but have not tested, that the onreset event handler isn't executed
either if you reset the form through JavaScript.

Finally, you can change the ACTION of a form if you want to:

document.forms[0].action = 'the_other_script.pl';

This can come in very handily if a form has to be submitted to another script in some
cases.

First of all, you should have a clear idea of what happens when a user submits a
form that has a JavaScript validation script:

1. The form is checked by a JavaScript like the one described below.
   If the script finds a mistake the submission is halted here. The user sees an alert
   and is asked to re-enter some data.
   
2. If nothing is wrong—or if JavaScript is disabled—the form is sent to the server and is processed by a CGI
   script.
   
3. If the CGI script finds a mistake it generates some HTML with an error message
   and sends it back to the user. In this case the user has to go back to the form,
   re-enter some values and again submit it.
   
4. If no mistakes are found, the CGI script does whatever it has to do with the data
   and directs the user to a Thank You page.
   

As you see, the form is checked for mistakes twice: by the JavaScript
and by the CGI script. The CGI check always works, since CGI is server side. The
JavaScript check only works when the user has JavaScript enabled.
It follows that the CGI check is the most reliable: it always works regardless of
what browser is used. Then why use a JavaScript check too?

The JavaScript check is very useful in addition to the CGI check because it
can catch mistakes before the form is actually sent to the server. Thus the user doesn't have to
use his back button to return to the form, something that may cause confusion, and then
search for the incorrect form field, which may cause even more confusion.
Therefore the JavaScript check is more user friendly than the CGI check.

In addition, when you use JavaScript the
server doesn't need to spend quite so much time in error handling and is thus a little
quicker. This only matters if you have lots and lots of forms, but it's good to
keep it in mind.

So JavaScript is not a fail-safe method of catching mistakes, but it is very
useful as an addition to CGI checks since it lightens the load on the server and is
more user friendly.
Therefore I recommend using both JavaScript form checks and CGI form checks. This
way, you get both user friendliness and security.

On this page I give the code you need for checking user input in a form.
Using these bits of code, you can write your own scripts to validate forms

I can't give you a complete script, because each form and each check are different.
You'll have to use the elements I explain on this page to build your own script.
I created an example form and script that you can study to
get the hang of it.

On this page I discuss the
limitations of using JavaScript to check a form, then I'll explain the
onsubmit event handler, followed by the few
methods and properties of the form itself.
Then it's time for
accessing the form elements and the specific syntax for accessing
the user defined
value of form elements.

See also Jeff Howden's excellent article
class="external" >Forms & JavaScript Living Together in Harmony
for some of the most common usability errors and their solutions.