Or do you want to send an email with a verification link, thus verifying that the user actually has access to the address? Or something inbetween (eg checking DNS)?

+ + on the rest of the post,
...but Fie! on your use of that wikipedia article as an athoritative source of legitimate ("trustworthy?") webmail providers because

1. webmail (meaning a mail service accessed via a browser) doesn't even enter OP's spec
     and
2. that list deals only with a few prominent providers, whereas many (most?) ISP's provide a webmails option in addition to POP3 and/or IMAP.

Note that I have no quarrel with the recommendation to "reject any suspicious/uncommon email providers." But defining either "suspicious" or "uncommon" with any particularity is going to be a major challenge, colored by personal experience and taste (sorta' like the Supreme Court Justice's comment re porn: "I know it when I see it."

webmail (meaning a mail service accessed via a browser) doesn't even enter OP's spec
yes, I know. That was my main point. the OP didn't appear to have even identified exactly what they wanted to do or what requirements they needed to work within.

But defining either "suspicious" or "uncommon" with any particularity is going to be a major challenge
Agreed. The wiki link was intended to get the OP thinking about such possibilities, not as a conclusive list for making that decision. (I do admit my post was a little ambiguous on that point. )

And if you're doing (a), then please don't reinvent the wheel as a square, but use Mail::RFC822::Address

indeed. And to clarify, I don't encourage reinventing something unless it will improve/replace it.

I deliberately tried to avoid discussing the 'how' details, and restricted the post to 'what' could be achieved, which I believe should always be done first.

Thanks to everyone for their response.

I'm working with a mature app, but it lacks email verification, therefore it opens itself up to account creation with an email address that doesn't exist.

Before the account is created, I would like to be able to, at a minimum, ping the email address to be sure it exists.

Pinging the existence of the domain name would not be sufficient when inappropriately using the domain name of popular free email address sites.

Hopefully, the email address pinging can distinguish between non-existent addresses versus full mailboxes.

Ideally, I would like to have email verification integrated at a later date.

Also, is there a downside to pinging the email address and waiting for a response before the software continues onto the account creation (or the denial thereof)?

Thanks.

I'm guessing that you don't mean "ping" to be taken literally in "ping the email address" because that's not what ping does. It may be feasible to ping the email server you're checking up on, but I've never heard of any way to ping an account ( - :-) - that admission may even prompt some wiser and more knowledgeable soul to post code to refute my belief!).

So maybe you should think about finger or another appropriate tool.

But,just as "cutting off the dog's tail by inches" is unnecessarily painful for the dog, I suspect verifying the existance of any particular email address -- EXCEPT BY requiring the owner to respond -- is going to be far more painful for you and your organization than "just doing it right in the first place;" i.e., "have email verification integrated" now.

As to downsides to the ping scheme, they include

1. delay
2. failure
       and, of course,
3. the inadequacy of that method of determining the validity of the email address used for account creation.