diff --git a/README.md b/README.md
index bfd1061..c933dbd 100644
--- a/README.md
+++ b/README.md
@@ -42,6 +42,7 @@ fn main() {
 }
 ```
 
+See the [examples](https://github.com/adumbidiot/lz-str-rs/tree/master/examples) directory for more examples.
 
 ## Testing
 ```bash
diff --git a/examples/binary-data.rs b/examples/binary-data.rs
new file mode 100644
index 0000000..04e4b72
--- /dev/null
+++ b/examples/binary-data.rs
@@ -0,0 +1,55 @@
+use std::convert::TryInto;
+
+// Sometimes, you may need to compress or decompress binary data.
+// This shows a way to handle that.
+fn main() {
+    let data = b"The quick brown fox jumps over the lazy dog";
+
+    // We pad every byte to make it a `u16`.
+    let wide_data: Vec<u16> = data.iter().copied().map(u16::from).collect();
+
+    let compressed_data = lz_str::compress(wide_data);
+
+    // Let's turn the output into bytes
+    // so we can show how to decompress from bytes.
+    // We make the data native endian.
+    let compressed_bytes: Vec<u8> = compressed_data
+        .iter()
+        .copied()
+        .flat_map(u16::to_ne_bytes)
+        .collect();
+
+    // We cannot decompress from an arbitrary string of bytes;
+    // those byte must be encoded wide characters (u16).
+
+    // First, let's ensure those bytes are a multiple of two,
+    // since two u8s form one u16.
+    // Make this a check instead of an `assert` in an actual application.
+    assert!(compressed_bytes.len() % 2 == 0);
+
+    // Now, let's form a `Vec<u16>` from the bytes.
+    // Here, we use u16::from_ne_bytes,
+    // since we made this byte array with native endian data.
+    // Make sure to use the correct function for your data.
+    // As an example, if you get your bytes from a network connection,
+    // you likely want to use `u16::from_be_bytes`.
+    let compressed_data: Vec<u16> = compressed_bytes
+        .chunks(2)
+        .map(|slice| {
+            // slice is always guaranteed to be 2 long here.
+            u16::from_ne_bytes(slice.try_into().unwrap())
+        })
+        .collect();
+
+    // Finally, call the decompress function.
+    // Again, make sure to do proper error checking in a real application.
+    let decompressed_data =
+        lz_str::decompress(compressed_data).expect("`compressed_bytes` is invalid");
+
+    // Remap the u16 to u8 via truncation.
+    // We do this in this example only to prove we get the same bytes back from the beginning.
+    let decompressed_data: Vec<u8> = decompressed_data.iter().copied().map(|v| v as u8).collect();
+
+    // We get the same bytes from the beginning.
+    assert!(decompressed_data == data);
+}