You cannot trust JavaScript at all. Not only can the viewer turn JavaScript off, but they can also change it. A JavaScript file is just like a web page: it gets downloaded to the client computer and then “displayed”. The viewer has full control over the file since it is on their computer, and malicious scripters can take advantage of this if you trust the “output” of JavaScript.
JavaScript should be used to assist the visitor, not to assist the server (or the webmaster). It should not be used to perform authentication, nor should it be used to calculate things that are then stored on the server. You would never want to calculate prices using JavaScript and then charge a customer based on that calculation, for example: you’ll end up giving some people a pretty hefty discount when they realize they can manipulate the calculation themselves. This does not mean that you cannot use JavaScript to show the viewer a quick summary of their prices or shopping cart total. It just means that you should, once they submit their purchase, calculate the prices on the server end as well. Remember that on the Internet, it only takes one person to take advantage of security flaws in your web site.
JavaScript is a great feature for making your web pages easier to use. But it should not be used in place of server-side programming. Your pages should always be accessible to people who do not use JavaScript (it is difficult to imagine what a JavaScript would look like through a translator or a voice reader, for example), and you should not rely on the JavaScript calculations coming to your end intact and trustworthy.