diff --git a/Control/Concurrent/ParallelIO/Local.hs b/Control/Concurrent/ParallelIO/Local.hs
index 8e8b26c..04d548c 100644
--- a/Control/Concurrent/ParallelIO/Local.hs
+++ b/Control/Concurrent/ParallelIO/Local.hs
@@ -21,7 +21,7 @@ module Control.Concurrent.ParallelIO.Local (
     parallelFirst, parallelFirstE,
 
     -- * Pool management
-    Pool, withPool, startPool, stopPool,
+    Pool, withPool, withBindingPool, startPool, startBindingPool, stopPool,
     extraWorkerWhileBlocked,
     
     -- * Advanced pool management
@@ -51,6 +51,7 @@ reflectExceptionsTo tid act = catchNonThreadKilled act (throwTo tid)
 -- | A thread pool, containing a maximum number of threads. The best way to
 -- construct one of these is using 'withPool'.
 data Pool = Pool {
+    pool_binding :: Bool,
     pool_threadcount :: Int,
     pool_sem :: QSem
   }
@@ -64,9 +65,16 @@ data Pool = Pool {
 --
 -- A better alternative is to see if you can use the 'withPool' variant.
 startPool :: Int -> IO Pool
-startPool threadcount
+startPool = startPool' False
+
+-- | 'startPool' binding threads to Haskell capabilities
+startBindingPool :: Int -> IO Pool
+startBindingPool = startPool' True
+
+startPool' :: Bool -> Int -> IO Pool
+startPool' binding threadcount
   | threadcount < 1 = error $ "startPool: thread count must be strictly positive (was " ++ show threadcount ++ ")"
-  | otherwise = fmap (Pool threadcount) $ newQSem (threadcount - 1)
+  | otherwise = fmap (Pool binding threadcount) $ newQSem (threadcount - 1)
 
 -- | Clean up a thread pool. If you don't call this from the main thread then no one holds the queue,
 -- the queue gets GC'd, the threads find themselves blocked indefinitely, and you get exceptions.
@@ -84,6 +92,9 @@ stopPool pool = replicateM_ (pool_threadcount pool - 1) $ killPoolWorkerFor pool
 withPool :: Int -> (Pool -> IO a) -> IO a
 withPool threadcount = bracket (startPool threadcount) stopPool
 
+-- | 'withPool' variant binding threads to Haskell capabilities.
+withBindingPool :: Int -> (Pool -> IO a) -> IO a
+withBindingPool threadcount = bracket (startBindingPool threadcount) stopPool
 
 -- | You should wrap any IO action used from your worker threads that may block with this method.
 -- It temporarily spawns another worker thread to make up for the loss of the old blocked
@@ -225,9 +236,12 @@ parallelE_ pool = fmap (map (either Just (\_ -> Nothing))) . parallelE pool
 parallel :: Pool -> [IO a] -> IO [a]
 parallel pool acts = mask $ \restore -> do
     main_tid <- myThreadId
-    resultvars <- forM acts $ \act -> do
+    nCaps <- getNumCapabilities
+    let enumeratedActs = zip (cycle [0..nCaps]) acts
+    resultvars <- forM enumeratedActs $ \(n, act) -> do
         resultvar <- newEmptyMVar
-        _tid <- forkIO $ bracket_ (killPoolWorkerFor pool) (spawnPoolWorkerFor pool) $ reflectExceptionsTo main_tid $ do
+        let fork = if pool_binding pool then forkOn n else forkIO
+        _tid <- fork $ bracket_ (killPoolWorkerFor pool) (spawnPoolWorkerFor pool) $ reflectExceptionsTo main_tid $ do
             res <- restore act
             -- Use tryPutMVar instead of putMVar so we get an exception if my brain has failed
             True <- tryPutMVar resultvar res