diff --git a/docs/getting_started.coffee.md b/docs/getting_started.coffee.md
index 9f8b64a53..d4d89d86b 100644
--- a/docs/getting_started.coffee.md
+++ b/docs/getting_started.coffee.md
@@ -117,6 +117,48 @@ For example:
         bottom: 50
         left: 72
         right: 72
+        
+## Switching to previous pages
+
+PDFKit normally flushes pages to the output file immediately when a new page is created, making
+it impossible to jump back and add content to previous pages.  This is normally not an issue, but
+in some circumstances it can be useful to add content to pages after the whole document, or a part
+of the document, has been created already.  Examples include adding page numbers, or filling in other
+parts of information you don't have until the rest of the document has been created.
+
+PDFKit has a `bufferPages` option in versions v0.7.0 and later that allows you to control when
+pages are flushed to the output file yourself rather than letting PDFKit handle that for you. To use
+it, just pass `bufferPages: true` as an option to the `PDFDocument` constructor.  Then, you can call
+`doc.switchToPage(pageNumber)` to switch to a previous page (page numbers start at 0).
+
+When you're ready to flush the buffered pages to the output file, call `flushPages`.
+This method is automatically called by `doc.end()`, so if you just want to buffer all pages in the document, you
+never need to call it.  Finally, there is a `bufferedPageRange` method, which returns the range
+of pages that are currently buffered.  Here is a small example that shows how you might add page 
+numbers to a document.
+
+    # create a document, and enable bufferPages mode
+    doc = new PDFDocument
+      bufferPages: true
+      
+    # add some content...
+    doc.addPage()
+    # ...
+    doc.addPage()
+    
+    # see the range of buffered pages
+    range = doc.bufferedPageRange() # => { start: 0, count: 2 }
+    
+    for i in [range.start...range.start + range.count]
+      doc.switchToPage(i)
+      doc.text "Page #{i + 1} of #{range.count}"
+        
+    # manually flush pages that have been buffered
+    doc.flushPages()
+    
+    # or, if you are at the end of the document anyway,
+    # doc.end() will call it for you automatically.
+    doc.end()
 
 ## Setting document metadata