It depends on what you are talking about, you could have a checklist for Concrete5, for the webserver and for SEO/analytics.

We just had our own checklist at my previous job, it included stuff like:
- Configure cache
- Disable error with stack trace output on page
- Check if url redirects work properly (and urls without www redirect to www)
- Are Google Analytics and other tracking codes added?
- Are the cron jobs configured for the automated jobs?
- Do the email notifications work (for example when someone submits a form or when you request a password reset)

This was just work a working document which continued to grow as we went along.

I'll add a few more:
- check site on different screen sizes and browsers
- enable pretty URL's
- check bulk SEO
- check external links (valid and open in new tab)
- check social links
- check pageload speeds
- add to search engines

- create a backup for this version

Thanks @Justin1978 and @WillemAnchor, that is helpful, I will create my own from those suggestions.
I was running my sites through the one of Chrome's apps 'Web Developer Checklist' but those things I mentioned above were failing. Just wondered if I should be preparing my sites better.
I've also noticed when validating through W3C there are some redactor issues that aren't being passed. Eg in a content block when you 'center' the text the style is being added but redactor is also adding which as far as the validator goes is 'bad'. See attached.

Also when an images is added through the content block, redactor is adding an ID of "image-marker' to each one and that is also being thrown up as 'bad' and it has duplicate ID's. See attached.

It's annoying and I'd like to know if there is a way around those things happening.

Hmmm, that rel-attribute seems like a bug to me. If it's redactor there's no workaround without mudling around in 3rd party Javascript. You could do a preg_replace on the content in the template if you know your way around regular expressions but that's nasty and only intended as a quick fix.

I would just report it as a bug in the bug tracker and ignore it. It's not valid html but it will still render properly and search engines will probably ignore it all together as it's irrelevant to them.

The duplicate id attribute is not pretty either by the way. An id has to be unique but your page will still render and work properly.

Thanks @Justin1978, I don't know my way around regular expressions so won't touch that.

I'll report both to bug tracker. That image ID is pretty much stopping one of my sites from validating with 28 instances of it!!