In terms of usability, breaking long scrolling pages into shorter pages is a good idea when you can so so logically. In other words, if you've got an article on "how to raise a puppy," you might break it into an introduction, a section on feeding the puppy, another section on paper-training it, another on teaching it to walk with a leash, etc. As a bonus, you'll have more opportunities to display page-related AdSense ads, and your pages will rank for a variety of "long-tail" searches.

There used to be a rule of thumb that said a page shouldn't scroll more than 2-1/2 screens. That may or may not be a good rule, but back in the late 1990s, an academic study showed that users preferred a long article on multiple pages over a somewhat shorter article on a single page. What's more, users thought the longer, multiple-page article was shorter than the scrolling article, possibly because clicking was easier than scrolling.

Hey Farmboy, I regularly break my articles up into pages, and my pageviews per visitor are still pretty good (in my humble opinion). I agree with what was said, if it's broken into logical parts, go for it.

On the otherhand, a healthy sized page can capture a good keyword in the serp...

will have fewer AdSense opportunities but possibly less likely to frustrate visitors.

Actually I think readers are less frustrated with short pages as long as they are broken up logically as others mentioned.

Reading online is different than reading a book. On a computer monitor a long dense page just looks overwhelming.

So breaking up a long article is good for the visitor as well as good for AdSense placement.

I break based on concept/theme/focus/keyword/voodoo/mojo.

Example:

If the article has sections: buying a widget, selling a widget, and maintaining a widget, then that's going to get at least three pages unless my research indicates that "widget" is a keyword trivial to dominate.

Most likely, my research will indicate that "widget" is non-trivial to dominate, and that it will behoove me to try to first get on page 1 for: "buying widget", "selling widget", and "widget maintenance".

Three pages means three times the intra-site links (which can be a very useful ranking booster). Three pages means Googlebot can decide this content has three different primary themes (if I put them all on one page, Googlebot may decide that there's just one primary theme: maintaining widgets).

Writing and dividing content is something best done while keyword/payout/traffic analysis for the relevant topic is displayed on the other computer screen.

Have you noticed though, that on a popular keyword, the dominator ussually has a large page.

A very large page CAN look good and not so intimidating if you make it flow well.

Use H tags, use CSS to make the H tags look good. Make your paragraphs relatively short, us H tags to define each section and make the whole page "skimable".

Also, use of pictures at strategic locations helps make it less intimidating.

If I see a LOOONNNGG page of nothing but text, I may not be interested in reading it, but if the bold H tags paint a good overall picture, I can use them to jump to parts I like. also the images give it a warmer, friendlier "read me" feel to it.

If you are after one hard to get keyword, I recommend a large page, but lots of effort and proof reading, layout design thinking, and re-thinking must be done. If your goal is several, not-so-hard to get keywords, then break it up into different pages.

Google's Webmaster Guidelines say to design sites for users, not search engines. That's worked well for me.

Yes, short paragraphs and subheadings throughout the page helps. Web users usually scan first to decide if a page of text is worth reading.

I'm not sure if I agree that long pages are more likely to be number one. On the harder to rank terms I think most of it has to do with how many links and the quality of the inbound links as well as (to a lesser extent) how many internal links.

When it's not a competitive search phrase it seems like anything goes in terms of page length.

I'm talking Google Search here. With the other searches it seems like on page factors may be weighted more.

Google's Webmaster Guidelines say to design sites for users, not search engines.

But breaking up the article into logical pages helps not only the user but also provides the SE with on topic content. These shorter pages can be found in the SE and bring you organic traffic which doesn't cost you anything except the effort of braking up the long page into smaller ones.

We regularly publish some very long articles, chiefly investigative reports. We usually break them into three or four pages, but even so the pages are still quite lengthy, sometimes 10 or more screens long.

We get very strong response to the stories and it really doesn't seem to matter whether the pages are long, kind of long or really, really long.

The only comments are the occasional ones from readers who complain that breaking an article into too many short pages makes it harder to print.

We actually had one guy who wrote indignantly that it was hard for him to print the longer articles in a format that would fit into his newsletter.

If you look to serve your visitors: provide both:
- provide a low bandwidth, low make-up one page version that covers it all (e.g. great for printing and reading off-line)
- provide a shorter more eye-candy containing versions that's split up over pages, contains many more illustrations but is a pain to print

That way your users get the choice.

Worry about SEA, ads, ... after your worry about serving the needs of your visitors and typically you'll be able to get the rest after that right just as well.

If you're going to serve both "printer-friendly" and Web-friendly pages, be sure that the printer-friendly versions are invisible to search crawlers unless you're willing to risk problems with "duplicate content" filters in search engines.

Printer-friendly versions of pages may also encourage content thieves, and sometimes they just aren't appropriate (or needed). For example, a page that makes extensive use of hyperlinks is less useful as a printed page than a newspaper-style story without links.

Have you noticed though, that on a popular keyword, the dominator ussually has a large page.

No, I definitely have not noticed that. I have noticed the opposite, but I suspect both cases are an illusory correlation -- we're more likely to notice anything out of the ordinary when looking at winning SERPs, and unlikely to remember how many times we've seen the same feature on losing SERPs.