diff --git a/Codec/Archive/Zip/Conduit/Types.hs b/Codec/Archive/Zip/Conduit/Types.hs
index 3582b81..705de3f 100644
--- a/Codec/Archive/Zip/Conduit/Types.hs
+++ b/Codec/Archive/Zip/Conduit/Types.hs
@@ -20,9 +20,9 @@ instance Exception ZipError where
 -- |(The beginning of) a single entry in a zip stream, which may be any file or directory.
 -- As per zip file conventions, directory names should end with a slash and have no data, but this library does not ensure that.
 data ZipEntry = ZipEntry
-  { zipEntryName :: ByteString -- ^File name, usually utf-8 encoded, with a trailing slash for directories
+  { zipEntryName :: ByteString -- ^File name (in posix format, no leading slashes), usually utf-8 encoded, with a trailing slash for directories
   , zipEntryTime :: LocalTime -- ^Modification time
-  , zipEntrySize :: Maybe Word64 -- ^Size of file data (if known, ignored on zip)
+  , zipEntrySize :: Maybe Word64 -- ^Size of file data (if known); checked on zipping and also used as hint to enable zip64
   }
 
 -- |Summary information at the end of a zip stream.
diff --git a/Codec/Archive/Zip/Conduit/UnZip.hs b/Codec/Archive/Zip/Conduit/UnZip.hs
index fa5e1ef..a574eda 100644
--- a/Codec/Archive/Zip/Conduit/UnZip.hs
+++ b/Codec/Archive/Zip/Conduit/UnZip.hs
@@ -86,7 +86,7 @@ fromDOSTime time date = LocalTime
     (fromIntegral $ time `shiftR` 5 .&. 0x3f)
     (fromIntegral $ time `shiftL` 1 .&. 0x3f))
 
--- |Stream a zip file, producing a sequence of entry headers and data blocks.
+-- |Stream process a zip file, producing a sequence of entry headers and data blocks.
 -- For example, this might produce: @Left (ZipEntry "directory\/" ...), Left (ZipEntry "directory\/file.txt" ...), Right "hello w", Right "orld!\\n", Left ...@
 -- The final result is summary information taken from the end of the zip file.
 -- No state is maintained during processing, and, in particular, any information in the central directory is discarded.
diff --git a/Codec/Archive/Zip/Conduit/Zip.hs b/Codec/Archive/Zip/Conduit/Zip.hs
index b17fa4b..169e519 100644
--- a/Codec/Archive/Zip/Conduit/Zip.hs
+++ b/Codec/Archive/Zip/Conduit/Zip.hs
@@ -1,3 +1,4 @@
+-- |Stream the creation of a zip file, e.g., as it's being uploaded.
 {-# LANGUAGE RecordWildCards #-}
 {-# LANGUAGE ViewPatterns #-}
 module Codec.Archive.Zip.Conduit.Zip
@@ -38,8 +39,9 @@ import           Data.Word (Word16, Word64)
 import           Codec.Archive.Zip.Conduit.Types
 import           Codec.Archive.Zip.Conduit.Internal
 
+-- |Options controlling zip file parameters and features
 data ZipOptions = ZipOptions
-  { zipOpt64 :: Bool -- ^Allow 'ZipDataSource's over 4GB (reduces compatibility in some cases)
+  { zipOpt64 :: Bool -- ^Allow 'ZipDataSource's over 4GB (reduces compatibility in some cases); this is automatically enabled for any files of known size (e.g., 'zipEntrySize')
   , zipOptCompressLevel :: Int -- ^Compress (0 = store only, 9 = best) zipped files (improves compatibility, since some unzip programs don't supported stored, streamed files, including the one in this package)
   , zipOptInfo :: ZipInfo -- ^Other parameters to store in the zip file
   }
@@ -58,15 +60,17 @@ infixr 7 ?*
 True ?* x = x
 False ?* _ = 0
 
+-- |The data contents for a 'ZipEntry'. For empty entries (e.g., directories), use 'mempty'.
 data ZipData m
-  = ZipDataByteString BSL.ByteString
-  | ZipDataSource (C.Source m BS.ByteString)
+  = ZipDataByteString BSL.ByteString -- ^A known ByteString, which will be fully evaluated (not streamed)
+  | ZipDataSource (C.Source m BS.ByteString) -- ^A byte stream producer, streamed (and compressed) directly into the zip
 
 instance Monad m => Monoid (ZipData m) where
   mempty = ZipDataByteString BSL.empty
   mappend (ZipDataByteString a) (ZipDataByteString b) = ZipDataByteString $ mappend a b
   mappend a b = ZipDataSource $ mappend (zipDataSource a) (zipDataSource b)
 
+-- |Use a file on disk as 'ZipData' (@'ZipDataSource' . 'CB.sourceFile'@).
 zipFileData :: MonadResource m => FilePath -> ZipData m
 zipFileData = ZipDataSource . CB.sourceFile
 
@@ -97,6 +101,11 @@ output = countBytes . sourcePut
 maxBound16 :: Integral n => n
 maxBound16 = fromIntegral (maxBound :: Word16)
 
+-- |Stream produce a zip file, reading a sequence of entries with data.
+-- Although file data is never kept in memory (beyond a single 'ZipDataByteString'), the format of zip files requires producing a final directory of entries at the end of the file, consuming an additional ~100 bytes of state per entry during streaming.
+-- The final result is the total size of the zip file.
+--
+-- Depending on options, the resulting zip file should be compatible with most unzipping applications.
 zipStream :: (MonadBase b m, PrimMonad b, MonadThrow m) => ZipOptions -> C.ConduitM (ZipEntry, ZipData m) BS.ByteString m Word64
 zipStream ZipOptions{..} = execStateC 0 $ do
   (cnt, cdir) <- next 0 (mempty :: P.Put)
@@ -120,10 +129,11 @@ zipStream ZipOptions{..} = execStateC 0 $ do
               +++ Z.compress) sdat -- level for Z.compress?
             , dataSize cdat)
           | otherwise = (left (fmap (id &&& fst)) sdat, usiz)
-        z64 = maybe zipOpt64 (maxBound32 <) (max <$> usiz <*> csiz)
+        z64 = maybe (zipOpt64 || any (maxBound32 <) zipEntrySize)
+          (maxBound32 <) (max <$> usiz <*> csiz)
         namelen = BS.length zipEntryName
         (time, date) = toDOSTime zipEntryTime
-        mcrc = either (const Nothing) (Just . crc32) cdat
+        mcrc = either (const Nothing) (Just . crc32) dat
     when (namelen > maxBound16) $ zipError $ BSC.unpack zipEntryName ++ ": entry name too long"
     let common = do
           P.putWord16le $ isLeft dat ?* bit 3
@@ -162,6 +172,7 @@ zipStream ZipOptions{..} = execStateC 0 $ do
         return r)
       (\b -> outsz $ ((fromJust usiz, fromJust mcrc), fromJust csiz) <$ CB.sourceLbs b)
       cdat
+    when (any (usz /=) zipEntrySize) $ zipError $ BSC.unpack zipEntryName ++ ": incorrect zipEntrySize"
     return $ do
       -- central directory
       let o64 = off >= maxBound32