Resource for journal acceptance rates
A few journals/publishers publish statistics about their acceptance rate, but this data is usually rather sparse, and not communicated on the journal webpage itself.
However, publishers often allow researchers in the fields of scientometrics, ethics or psychology access to their (anonymized) database of paper review/acceptance/publication timelines. You can find papers such as this one (and references therein), which then analyze the raw data and publish some statistics about it. We learn, for example, that the acceptance ratio for Physical Review Letters in 2012 was at 32%.
Other nuggets of information one can gleam include:
in the Europhysics Letters study Michael Schreiber expressed concern about comparatively high acceptance rates for July-submitted papers
which is always good to know if you happen to work in that field :)
There are a few studies on rejection rates in specific areas such as Atmospheric Sciences and Ecology. In a seminar talk given by Elsevier at my University Library, the indication was that 40-90% rejection rates were not unusual. Typically, high profile journals (usually identified by a high Impact factors) will be at the upper end of the range. I edit a ISI listed journal and we consistently have about a 50% rejection rates. I believe the norm for many fields is in this neighbourhood.
If you are looking for a journal with an exactly known constant rejection rate for a paper you should go for the JUR.
On the Elsevier journal finder service website you can find some data about the acceptance rate, but only for journals published by Elsevier (of course). You have to fill the name of your article and add the abstract, than you will find the acceptance rates between the results (together with Impact factor, acceptance time etc.). Unfortunately, you can´t just browse among journals, you will see only the proposed journals which (maybe) suits to you.