diff --git a/README.md b/README.md
index 1cddfbfd..fe79bf74 100644
--- a/README.md
+++ b/README.md
@@ -59,12 +59,13 @@ pg.connect(function(err, client, done) {
 
 Creates an instance of a query cursor.  Pass this instance to node-postgres [`client#query`](https://github.com/brianc/node-postgres/wiki/Client#wiki-method-query-parameterized)
 
-#### cursor#read(int rowCount, function callback(Error err, Array rows)
+#### cursor#read(int rowCount, function callback(Error err, Array rows, Result result)
 
 Read `rowCount` rows from the cursor instance.  The `callback` will be called when the rows are available, loaded into memory, parsed, and converted to JavaScript types.
 
 If the cursor has read to the end of the result sets all subsequent calls to `cursor#read` will return a 0 length array of rows.  I'm open to other ways to signal the end of a cursor, but this has worked out well for me so far.
 
+`result` is a special [https://github.com/brianc/node-postgres/wiki/Query#result-object](Result) object that can be used to accumulate rows.
 
 #### cursor#close(function callback(Error err))
 
diff --git a/index.js b/index.js
index 70a90f57..71c12522 100644
--- a/index.js
+++ b/index.js
@@ -70,7 +70,8 @@ Cursor.prototype._sendRows = function() {
     //within the call to this callback
     this._cb = null
     if(cb) {
-      cb(null, this._rows)
+      this._result.rows = this._rows
+      cb(null, this._rows, this._result)
     }
     this._rows = []
   }.bind(this))
diff --git a/test/index.js b/test/index.js
index 8f04ccc2..af4f041c 100644
--- a/test/index.js
+++ b/test/index.js
@@ -116,4 +116,15 @@ describe('cursor', function() {
       })
     })
   })
+
+  it('returns result along with rows', function(done) {
+    var cursor = this.pgCursor(text)
+    cursor.read(1, function(err, rows, result) {
+      assert.ifError(err)
+      assert.equal(rows.length, 1)
+      assert.strictEqual(rows, result.rows)
+      assert.deepEqual(result.fields.map(f => f.name), ['num'])
+      done()
+    })
+  })
 })