diff --git a/CHANGELOG b/CHANGELOG
--- a/CHANGELOG
+++ b/CHANGELOG
@@ -1,5 +1,9 @@
 # Changelog
 
+- 0.1.2 (2025-10-21)
+  * Improves padding handling for the strict bytestring case, yielding
+    performance increases for both 'hash' and 'hmac'.
+
 - 0.1.1 (2025-01-07)
   * Minor bytestring-handling optimisation.
 
diff --git a/lib/Crypto/Hash/RIPEMD160.hs b/lib/Crypto/Hash/RIPEMD160.hs
--- a/lib/Crypto/Hash/RIPEMD160.hs
+++ b/lib/Crypto/Hash/RIPEMD160.hs
@@ -113,15 +113,85 @@
 -- k such that (l + 1 + k) mod 64 = 56
 sol :: Word64 -> Word64
 sol l =
-  let r = 56 - fi l `mod` 64 - 1 :: Integer -- fi prevents underflow
+  let r = 56 - fi l `rem` 64 - 1 :: Integer -- fi prevents underflow
   in  fi (if r < 0 then r + 64 else r)
 
 pad :: BS.ByteString -> BS.ByteString
-pad m@(BI.PS _ _ (fi -> l)) = BL.toStrict . BSB.toLazyByteString $ padded where
-  padded = BSB.byteString m <> fill (sol l) (BSB.word8 0x80)
-  fill j !acc
-    | j == 0 = acc <> BSB.word64LE (l * 8)
-    | otherwise = fill (pred j) (acc <> BSB.word8 0x00)
+pad m@(BI.PS _ _ (fi -> l)) =
+    BL.toStrict . BE.toLazyByteStringWith
+      (BE.safeStrategy 128 BE.smallChunkSize) mempty $ padded
+  where
+    padded = BSB.byteString m
+          <> fill (sol l) (BSB.word8 0x80)
+          <> BSB.word64LE (l * 8)
+
+    fill j !acc
+      | j `rem` 8 == 0 =
+             loop64 j acc
+      | (j - 7) `rem` 8 == 0 =
+             loop64 (j - 7) acc
+          <> BSB.word32LE 0x00
+          <> BSB.word16LE 0x00
+          <> BSB.word8 0x00
+      | (j - 6) `rem` 8 == 0 =
+             loop64 (j - 6) acc
+          <> BSB.word32LE 0x00
+          <> BSB.word16LE 0x00
+      | (j - 5) `rem` 8 == 0 =
+             loop64 (j - 5) acc
+          <> BSB.word32LE 0x00
+          <> BSB.word8 0x00
+      | (j - 4) `rem` 8 == 0 =
+             loop64 (j - 4) acc
+          <> BSB.word32LE 0x00
+      | (j - 3) `rem` 8 == 0 =
+             loop64 (j - 3) acc
+          <> BSB.word16LE 0x00
+          <> BSB.word8 0x00
+      | (j - 2) `rem` 8 == 0 =
+             loop64 (j - 2) acc
+          <> BSB.word16LE 0x00
+      | (j - 1) `rem` 8 == 0 =
+             loop64 (j - 1) acc
+          <> BSB.word8 0x00
+
+      | j `rem` 4 == 0 =
+             loop32 j acc
+      | (j - 3) `rem` 4 == 0 =
+             loop32 (j - 3) acc
+          <> BSB.word16LE 0x00
+          <> BSB.word8 0x00
+      | (j - 2) `rem` 4 == 0 =
+             loop32 (j - 2) acc
+          <> BSB.word16LE 0x00
+      | (j - 1) `rem` 4 == 0 =
+             loop32 (j - 1) acc
+          <> BSB.word8 0x00
+
+      | j `rem` 2 == 0 =
+             loop16 j acc
+      | (j - 1) `rem` 2 == 0 =
+             loop16 (j - 1) acc
+          <> BSB.word8 0x00
+
+      | otherwise =
+            loop8 j acc
+
+    loop64 j !acc
+      | j == 0 = acc
+      | otherwise = loop64 (j - 8) (acc <> BSB.word64LE 0x00)
+
+    loop32 j !acc
+      | j == 0 = acc
+      | otherwise = loop32 (j - 4) (acc <> BSB.word32LE 0x00)
+
+    loop16 j !acc
+      | j == 0 = acc
+      | otherwise = loop16 (j - 2) (acc <> BSB.word16LE 0x00)
+
+    loop8 j !acc
+      | j == 0 = acc
+      | otherwise = loop8 (pred j) (acc <> BSB.word8 0x00)
 
 pad_lazy :: BL.ByteString -> BL.ByteString
 pad_lazy (BL.toChunks -> m) = BL.fromChunks (walk 0 m) where
diff --git a/ppad-ripemd160.cabal b/ppad-ripemd160.cabal
--- a/ppad-ripemd160.cabal
+++ b/ppad-ripemd160.cabal
@@ -1,6 +1,6 @@
 cabal-version:      3.0
 name:               ppad-ripemd160
-version:            0.1.1
+version:            0.1.2
 synopsis:           The RIPEMD-160 hashing algorithm.
 license:            MIT
 license-file:       LICENSE