diff --git a/README.md b/README.md
index 350f1f3..36a6655 100644
--- a/README.md
+++ b/README.md
@@ -27,7 +27,7 @@ maintained with the occasional fix for edge cases and depends on Tokio v0.2.
 
 `FixedInt` casts integers to bytes by either copying the underlying memory or
 performing a transmutation. The encoded values use machine endianness
-(little-endian on x86).
+(little-endian on x86, big-endian where applicable).
 
 ## VarInt
 
diff --git a/src/fixed.rs b/src/fixed.rs
index fd0a24d..1f0eb33 100644
--- a/src/fixed.rs
+++ b/src/fixed.rs
@@ -49,12 +49,37 @@ macro_rules! impl_fixedint {
 
             fn encode_fixed(self, dst: &mut [u8]) {
                 assert_eq!(dst.len(), Self::REQUIRED_SPACE);
-                let encoded = unsafe { &*(&self as *const $t as *const [u8; $sz]) };
+
+                #[allow(unused_mut)]
+                let mut encoded = unsafe { &*(&self as *const $t as *const [u8; $sz]) };
+
+                #[cfg(target_endian = "big")]
+                {
+                    let mut encoded_rev = [0 as u8; Self::REQUIRED_SPACE];
+                    encoded_rev.copy_from_slice(encoded);
+                    encoded_rev.reverse();
+                    dst.clone_from_slice(&encoded_rev);
+                    return;
+                }
+
                 dst.clone_from_slice(encoded);
             }
+
             fn decode_fixed(src: &[u8]) -> $t {
                 assert_eq!(src.len(), Self::REQUIRED_SPACE);
-                return unsafe { (src.as_ptr() as *const $t).read_unaligned() };
+                #[cfg(target_endian = "little")]
+                let src_fin = src;
+
+                #[cfg(target_endian = "big")]
+                let mut src_fin = [0 as u8; Self::REQUIRED_SPACE];
+
+                #[cfg(target_endian = "big")]
+                {
+                    src_fin.copy_from_slice(src);
+                    src_fin.reverse();
+                }
+
+                return unsafe { (src_fin.as_ptr() as *const $t).read_unaligned() };
             }
         }
     };
diff --git a/src/fixed_tests.rs b/src/fixed_tests.rs
index b2d75c3..83af0b4 100644
--- a/src/fixed_tests.rs
+++ b/src/fixed_tests.rs
@@ -13,21 +13,33 @@ mod tests {
     #[test]
     fn test_u32_enc() {
         let result = (32 as u32).encode_fixed_vec();
+        #[cfg(target_endian = "big")]
+        assert_eq!(result, vec![0, 0, 0, 32]);
+        #[cfg(target_endian = "little")]
         assert_eq!(result, vec![32, 0, 0, 0]);
     }
     #[test]
     fn test_u16_enc() {
         let result = (256 as u16).encode_fixed_vec();
+        #[cfg(target_endian = "big")]
+        assert_eq!(result, vec![1, 0]);
+        #[cfg(target_endian = "little")]
         assert_eq!(result, vec![0, 1]);
     }
     #[test]
     fn test_i16_enc() {
         let result = (-32768 as i16).encode_fixed_vec();
+        #[cfg(target_endian = "big")]
+        assert_eq!(result, vec![128, 0]);
+        #[cfg(target_endian = "little")]
         assert_eq!(result, vec![0, 128]);
     }
     #[test]
     fn test_i32_enc() {
         let result = (-32767 as i32).encode_fixed_vec();
+        #[cfg(target_endian = "big")]
+        assert_eq!(result, vec![255, 255, 128, 1]);
+        #[cfg(target_endian = "little")]
         assert_eq!(result, vec![1, 128, 255, 255]);
     }
 
@@ -45,6 +57,9 @@ mod tests {
     fn test_i32_enc_light() {
         let int = -32767 as i32;
         let result = int.encode_fixed_light();
+        #[cfg(target_endian = "big")]
+        assert_eq!(result, &[255, 255, 128, 1]);
+        #[cfg(target_endian = "little")]
         assert_eq!(result, &[1, 128, 255, 255]);
     }
     #[test]
diff --git a/src/reader.rs b/src/reader.rs
index aaee8f0..8910783 100644
--- a/src/reader.rs
+++ b/src/reader.rs
@@ -5,7 +5,7 @@ use crate::fixed::FixedInt;
 use crate::varint::{VarInt, VarIntMaxSize, MSB};
 
 #[cfg(feature = "tokio_async")]
-use tokio::io::{AsyncReadExt, AsyncRead};
+use tokio::io::{AsyncRead, AsyncReadExt};
 
 #[cfg(feature = "futures_async")]
 use futures_util::{io::AsyncRead, io::AsyncReadExt};
@@ -40,7 +40,10 @@ pub struct VarIntProcessor {
 
 impl VarIntProcessor {
     fn new<VI: VarIntMaxSize>() -> VarIntProcessor {
-        VarIntProcessor { maxsize: VI::varint_max_size(), ..VarIntProcessor::default() }
+        VarIntProcessor {
+            maxsize: VI::varint_max_size(),
+            ..VarIntProcessor::default()
+        }
     }
     fn push(&mut self, b: u8) -> Result<()> {
         if self.i >= self.maxsize {
diff --git a/src/varint.rs b/src/varint.rs
index 20069fc..0fd319a 100644
--- a/src/varint.rs
+++ b/src/varint.rs
@@ -72,7 +72,7 @@ pub(crate) trait VarIntMaxSize {
 
 impl<VI: VarInt> VarIntMaxSize for VI {
     fn varint_max_size() -> usize {
-        (size_of::<VI>()*8+7)/7
+        (size_of::<VI>() * 8 + 7) / 7
     }
 }
 
diff --git a/src/varint_tests.rs b/src/varint_tests.rs
index 432e2a0..ba70e86 100644
--- a/src/varint_tests.rs
+++ b/src/varint_tests.rs
@@ -204,6 +204,9 @@ mod tests {
     #[test]
     fn test_regression_22() {
         let mut encoded: Vec<u8> = (0x112233 as u64).encode_var_vec();
-        assert_eq!(encoded.as_slice().read_varint::<i8>().unwrap_err().kind(), std::io::ErrorKind::InvalidData);
+        assert_eq!(
+            encoded.as_slice().read_varint::<i8>().unwrap_err().kind(),
+            std::io::ErrorKind::InvalidData
+        );
     }
 }