diff --git a/ethereum-types/src/hash.rs b/ethereum-types/src/hash.rs
index 82372ee64..070124268 100644
--- a/ethereum-types/src/hash.rs
+++ b/ethereum-types/src/hash.rs
@@ -74,12 +74,12 @@ macro_rules! impl_uint_conversions {
 
 			fn from_uint(value: &$uint) -> Self {
 				let mut ret = $hash::zero();
-				value.to_big_endian(ret.as_bytes_mut());
+				value.write_as_big_endian(ret.as_bytes_mut());
 				ret
 			}
 
 			fn into_uint(&self) -> $uint {
-				$uint::from(self.as_ref() as &[u8])
+				$uint::from_big_endian(self.as_ref() as &[u8])
 			}
 		}
 	};
diff --git a/ethereum-types/src/uint.rs b/ethereum-types/src/uint.rs
index 7b9b8f07e..5dfbdb310 100644
--- a/ethereum-types/src/uint.rs
+++ b/ethereum-types/src/uint.rs
@@ -102,8 +102,8 @@ mod tests {
 	#[test]
 	fn fixed_arrays_roundtrip() {
 		let raw: U256 = "7094875209347850239487502394881".into();
-		let array: [u8; 32] = raw.into();
-		let new_raw = array.into();
+		let array: [u8; 32] = raw.to_big_endian();
+		let new_raw = U256::from_big_endian(&array);
 
 		assert_eq!(raw, new_raw);
 	}
diff --git a/primitive-types/impls/codec/src/lib.rs b/primitive-types/impls/codec/src/lib.rs
index feacec08a..905d17d12 100644
--- a/primitive-types/impls/codec/src/lib.rs
+++ b/primitive-types/impls/codec/src/lib.rs
@@ -19,8 +19,7 @@ macro_rules! impl_uint_codec {
 	($name: ident, $len: expr) => {
 		impl $crate::codec::Encode for $name {
 			fn using_encoded<R, F: FnOnce(&[u8]) -> R>(&self, f: F) -> R {
-				let mut bytes = [0u8; $len * 8];
-				self.to_little_endian(&mut bytes);
+				let bytes = self.to_little_endian();
 				bytes.using_encoded(f)
 			}
 		}
diff --git a/primitive-types/impls/rlp/src/lib.rs b/primitive-types/impls/rlp/src/lib.rs
index 71382a303..9b17ea4f7 100644
--- a/primitive-types/impls/rlp/src/lib.rs
+++ b/primitive-types/impls/rlp/src/lib.rs
@@ -23,8 +23,7 @@ macro_rules! impl_uint_rlp {
 		impl $crate::rlp::Encodable for $name {
 			fn rlp_append(&self, s: &mut $crate::rlp::RlpStream) {
 				let leading_empty_bytes = $size * 8 - (self.bits() + 7) / 8;
-				let mut buffer = [0u8; $size * 8];
-				self.to_big_endian(&mut buffer);
+				let buffer = self.to_big_endian();
 				s.encoder().encode_value(&buffer[leading_empty_bytes..]);
 			}
 		}
@@ -35,7 +34,7 @@ macro_rules! impl_uint_rlp {
 					if !bytes.is_empty() && bytes[0] == 0 {
 						Err($crate::rlp::DecoderError::RlpInvalidIndirection)
 					} else if bytes.len() <= $size * 8 {
-						Ok($name::from(bytes))
+						Ok($name::from_big_endian(bytes))
 					} else {
 						Err($crate::rlp::DecoderError::RlpIsTooBig)
 					}
diff --git a/primitive-types/impls/serde/src/lib.rs b/primitive-types/impls/serde/src/lib.rs
index 63fe535cb..50587a91d 100644
--- a/primitive-types/impls/serde/src/lib.rs
+++ b/primitive-types/impls/serde/src/lib.rs
@@ -32,8 +32,7 @@ macro_rules! impl_uint_serde {
 				S: $crate::serde::Serializer,
 			{
 				let mut slice = [0u8; 2 + 2 * $len * 8];
-				let mut bytes = [0u8; $len * 8];
-				self.to_big_endian(&mut bytes);
+				let bytes = self.to_big_endian();
 				$crate::serialize::serialize_uint(&mut slice, &bytes, serializer)
 			}
 		}
@@ -48,7 +47,7 @@ macro_rules! impl_uint_serde {
 					deserializer,
 					$crate::serialize::ExpectedLen::Between(0, &mut bytes),
 				)?;
-				Ok(bytes[0..wrote].into())
+				Ok(Self::from_big_endian(&bytes[0..wrote]))
 			}
 		}
 	};
diff --git a/rlp/tests/tests.rs b/rlp/tests/tests.rs
index 768de3dd2..c5224fd5d 100644
--- a/rlp/tests/tests.rs
+++ b/rlp/tests/tests.rs
@@ -226,7 +226,7 @@ fn encode_u256() {
 		ETestPair::from((U256::from(0x0100_0000_u64), hex!("8401000000"))),
 		ETestPair::from((U256::from(0xffff_ffff_u64), hex!("84ffffffff"))),
 		ETestPair::from((
-			hex!("  8090a0b0c0d0e0f00910203040506077000000000000000100000000000012f0").into(),
+			U256::from_big_endian(&hex!("  8090a0b0c0d0e0f00910203040506077000000000000000100000000000012f0")),
 			hex!("a08090a0b0c0d0e0f00910203040506077000000000000000100000000000012f0"),
 		)),
 	];
@@ -482,7 +482,7 @@ fn decode_untrusted_u256() {
 		DTestPair::from((U256::from(0x0100_0000_u64), hex!("8401000000"))),
 		DTestPair::from((U256::from(0xffff_ffff_u64), hex!("84ffffffff"))),
 		DTestPair::from((
-			hex!("  8090a0b0c0d0e0f00910203040506077000000000000000100000000000012f0").into(),
+			U256::from_big_endian(&hex!("  8090a0b0c0d0e0f00910203040506077000000000000000100000000000012f0")),
 			hex!("a08090a0b0c0d0e0f00910203040506077000000000000000100000000000012f0"),
 		)),
 	];
diff --git a/uint/benches/bigint.rs b/uint/benches/bigint.rs
index c092dfc5f..697f641d4 100644
--- a/uint/benches/bigint.rs
+++ b/uint/benches/bigint.rs
@@ -78,8 +78,7 @@ criterion_group!(
 criterion_main!(bigint);
 
 fn to_biguint(x: U256) -> BigUint {
-	let mut bytes = [0u8; 32];
-	x.to_little_endian(&mut bytes);
+	let mut bytes = x.to_little_endian();
 	BigUint::from_bytes_le(&bytes)
 }
 
@@ -662,8 +661,8 @@ fn from_fixed_array(c: &mut Criterion) {
 		[255, 0, 0, 123, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 121, 0, 0, 0, 0, 0, 213, 0, 0, 0, 0, 0, 0];
 	c.bench_function("from_fixed_array", move |b| {
 		b.iter(|| {
-			let _: U512 = black_box(black_box(ary512).into());
-			let _: U256 = black_box(black_box(ary256).into());
+			let _: U512 = black_box(U512::from_big_endian(black_box(&ary512)));
+			let _: U256 = black_box(U256::from_big_endian(black_box(&ary256)));
 		})
 	});
 }
diff --git a/uint/src/uint.rs b/uint/src/uint.rs
index bc4b7416c..08081412a 100644
--- a/uint/src/uint.rs
+++ b/uint/src/uint.rs
@@ -743,9 +743,17 @@ macro_rules! construct_uint {
 				(arr[index / 8] >> (((index % 8)) * 8)) as u8
 			}
 
+			/// convert to big-endian bytes.
+			#[inline]
+			pub fn to_big_endian(&self)  -> [u8; $n_words * 8] {
+				let mut bytes = [0u8; $n_words * 8];
+				self.write_as_big_endian(&mut bytes);
+				bytes
+			}
+
 			/// Write to the slice in big-endian format.
 			#[inline]
-			pub fn to_big_endian(&self, bytes: &mut [u8]) {
+			pub fn write_as_big_endian(&self, bytes: &mut [u8]) {
 				use $crate::byteorder::{ByteOrder, BigEndian};
 				debug_assert!($n_words * 8 == bytes.len());
 				for i in 0..$n_words {
@@ -753,9 +761,16 @@ macro_rules! construct_uint {
 				}
 			}
 
-			/// Write to the slice in little-endian format.
+			/// convert to little-endian bytes.
 			#[inline]
-			pub fn to_little_endian(&self, bytes: &mut [u8]) {
+			pub fn to_little_endian(&self) -> [u8; $n_words * 8] {
+				let mut bytes = [0u8; $n_words * 8];
+				self.write_as_little_endian(&mut bytes);
+				bytes
+			}
+
+			#[inline]
+			pub fn write_as_little_endian(&self, bytes: &mut [u8]) {
 				use $crate::byteorder::{ByteOrder, LittleEndian};
 				debug_assert!($n_words * 8 == bytes.len());
 				for i in 0..$n_words {
@@ -1307,26 +1322,6 @@ macro_rules! construct_uint {
 			}
 		}
 
-		impl $crate::core_::convert::From<$name> for [u8; $n_words * 8] {
-			fn from(number: $name) -> Self {
-				let mut arr = [0u8; $n_words * 8];
-				number.to_big_endian(&mut arr);
-				arr
-			}
-		}
-
-		impl $crate::core_::convert::From<[u8; $n_words * 8]> for $name {
-			fn from(bytes: [u8; $n_words * 8]) -> Self {
-				Self::from(&bytes)
-			}
-		}
-
-		impl<'a> $crate::core_::convert::From<&'a [u8; $n_words * 8]> for $name {
-			fn from(bytes: &[u8; $n_words * 8]) -> Self {
-				Self::from(&bytes[..])
-			}
-		}
-
 		impl $crate::core_::default::Default for $name {
 			fn default() -> Self {
 				$name::zero()
@@ -1360,13 +1355,6 @@ macro_rules! construct_uint {
 		$crate::impl_map_from!($name, i32, i64);
 		$crate::impl_map_from!($name, isize, i64);
 
-		// Converts from big endian representation.
-		impl<'a> $crate::core_::convert::From<&'a [u8]> for $name {
-			fn from(bytes: &[u8]) -> $name {
-				Self::from_big_endian(bytes)
-			}
-		}
-
 		$crate::impl_try_from_for_primitive!($name, u8);
 		$crate::impl_try_from_for_primitive!($name, u16);
 		$crate::impl_try_from_for_primitive!($name, u32);
@@ -1736,8 +1724,7 @@ macro_rules! construct_uint {
 					$crate::hex::decode_to_slice(encoded, out).map_err(Self::Err::from)?;
 				}
 
-				let bytes_ref: &[u8] = &bytes;
-				Ok(From::from(bytes_ref))
+				Ok(Self::from_big_endian(&bytes))
 			}
 		}
 
@@ -1787,7 +1774,7 @@ macro_rules! impl_quickcheck_arbitrary_for_uint {
 					}
 				});
 
-				res.as_ref().into()
+				Self::from_big_endian(res.as_ref())
 			}
 		}
 	};
@@ -1809,7 +1796,7 @@ macro_rules! impl_arbitrary_for_uint {
 			fn arbitrary(u: &mut $crate::arbitrary::Unstructured<'_>) -> $crate::arbitrary::Result<Self> {
 				let mut res = [0u8; $n_bytes];
 				u.fill_buffer(&mut res)?;
-				Ok(Self::from(res))
+				Ok(Self::from_big_endian(&res))
 			}
 		}
 	};
diff --git a/uint/tests/uint_tests.rs b/uint/tests/uint_tests.rs
index a830e488e..61c03c86d 100644
--- a/uint/tests/uint_tests.rs
+++ b/uint/tests/uint_tests.rs
@@ -175,18 +175,18 @@ fn uint256_from() {
 	assert_eq!(e, ud);
 
 	// test initialization from bytes
-	let va = U256::from(&[10u8][..]);
+	let va = U256::from_big_endian(&[10u8][..]);
 	assert_eq!(e, va);
 
 	// more tests for initialization from bytes
-	assert_eq!(U256([0x1010, 0, 0, 0]), U256::from(&[0x10u8, 0x10][..]));
-	assert_eq!(U256([0x12f0, 0, 0, 0]), U256::from(&[0x12u8, 0xf0][..]));
-	assert_eq!(U256([0x12f0, 0, 0, 0]), U256::from(&[0, 0x12u8, 0xf0][..]));
-	assert_eq!(U256([0x12f0, 0, 0, 0]), U256::from(&[0, 0, 0, 0, 0, 0, 0, 0x12u8, 0xf0][..]));
-	assert_eq!(U256([0x12f0, 1, 0, 0]), U256::from(&[1, 0, 0, 0, 0, 0, 0, 0x12u8, 0xf0][..]));
+	assert_eq!(U256([0x1010, 0, 0, 0]), U256::from_big_endian(&[0x10u8, 0x10][..]));
+	assert_eq!(U256([0x12f0, 0, 0, 0]), U256::from_big_endian(&[0x12u8, 0xf0][..]));
+	assert_eq!(U256([0x12f0, 0, 0, 0]), U256::from_big_endian(&[0, 0x12u8, 0xf0][..]));
+	assert_eq!(U256([0x12f0, 0, 0, 0]), U256::from_big_endian(&[0, 0, 0, 0, 0, 0, 0, 0x12u8, 0xf0][..]));
+	assert_eq!(U256([0x12f0, 1, 0, 0]), U256::from_big_endian(&[1, 0, 0, 0, 0, 0, 0, 0x12u8, 0xf0][..]));
 	assert_eq!(
 		U256([0x12f0, 1, 0x0910203040506077, 0x8090a0b0c0d0e0f0]),
-		U256::from(
+		U256::from_big_endian(
 			&[
 				0x80, 0x90, 0xa0, 0xb0, 0xc0, 0xd0, 0xe0, 0xf0, 0x09, 0x10, 0x20, 0x30, 0x40, 0x50, 0x60, 0x77, 0, 0,
 				0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0x12u8, 0xf0
@@ -195,7 +195,7 @@ fn uint256_from() {
 	);
 	assert_eq!(
 		U256([0x00192437100019fa, 0x243710, 0, 0]),
-		U256::from(&[0x24u8, 0x37, 0x10, 0, 0x19, 0x24, 0x37, 0x10, 0, 0x19, 0xfa][..])
+		U256::from_big_endian(&[0x24u8, 0x37, 0x10, 0, 0x19, 0x24, 0x37, 0x10, 0, 0x19, 0xfa][..])
 	);
 
 	// test initializtion from string
@@ -273,9 +273,8 @@ fn uint256_try_into_primitives() {
 fn uint256_to() {
 	let hex = "8090a0b0c0d0e0f00910203040506077583a2cf8264910e1436bda32571012f0";
 	let uint = U256::from_str(hex).unwrap();
-	let mut bytes = [0u8; 32];
-	uint.to_big_endian(&mut bytes);
-	let uint2 = U256::from(&bytes[..]);
+	let bytes = uint.to_big_endian();
+	let uint2 = U256::from_big_endian(&bytes[..]);
 	assert_eq!(uint, uint2);
 }
 
@@ -893,7 +892,7 @@ fn big_endian() {
 
 	assert_eq!(source, U256::from(1));
 
-	source.to_big_endian(&mut target);
+	source.write_as_big_endian(&mut target);
 	assert_eq!(
 		vec![
 			0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8,
@@ -905,7 +904,7 @@ fn big_endian() {
 	let source = U256([512, 0, 0, 0]);
 	let mut target = vec![0u8; 32];
 
-	source.to_big_endian(&mut target);
+	source.write_as_big_endian(&mut target);
 	assert_eq!(
 		vec![
 			0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8,
@@ -917,7 +916,7 @@ fn big_endian() {
 	let source = U256([0, 512, 0, 0]);
 	let mut target = vec![0u8; 32];
 
-	source.to_big_endian(&mut target);
+	source.write_as_big_endian(&mut target);
 	assert_eq!(
 		vec![
 			0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8, 0u8,
@@ -927,7 +926,7 @@ fn big_endian() {
 	);
 
 	let source = U256::from_str("0102030405060708090a0b0c0d0e0f101112131415161718191a1b1c1d1e1f20").unwrap();
-	source.to_big_endian(&mut target);
+	source.write_as_big_endian(&mut target);
 	assert_eq!(
 		vec![
 			0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f, 0x10, 0x11, 0x12,
@@ -1008,7 +1007,7 @@ fn little_endian() {
 		0xbb, 0xc5, 0x3c, 0x7d, 0x2b, 0x72, 0xe5, 0xf6, 0xa3, 0x1d, 0xca, 0x2c, 0x02, 0x00,
 	];
 	let mut result = [0u8; 32];
-	number.to_little_endian(&mut result);
+	number.write_as_little_endian(&mut result);
 	assert_eq!(expected, result);
 }
 
@@ -1019,11 +1018,11 @@ fn slice_roundtrip() {
 		107, 109, 113, 127,
 	];
 
-	let u256: U256 = (&raw[..]).into();
+	let u256 = U256::from_big_endian(&raw[..]);
 
 	let mut new_raw = [0u8; 32];
 
-	u256.to_big_endian(&mut new_raw);
+	u256.write_as_big_endian(&mut new_raw);
 
 	assert_eq!(&raw, &new_raw);
 }
@@ -1039,7 +1038,7 @@ fn slice_roundtrip_le() {
 
 	let mut new_raw = [0u8; 32];
 
-	u256.to_little_endian(&mut new_raw);
+	u256.write_as_little_endian(&mut new_raw);
 
 	assert_eq!(&raw, &new_raw);
 }
@@ -1055,7 +1054,7 @@ fn slice_roundtrip_le2() {
 
 	let mut new_raw = [0u8; 32];
 
-	u256.to_little_endian(&mut new_raw);
+	u256.write_as_little_endian(&mut new_raw);
 
 	assert_eq!(&raw, &new_raw[..31]);
 }
@@ -1090,17 +1089,17 @@ fn from_big_endian() {
 fn into_fixed_array() {
 	let expected: [u8; 32] =
 		[0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1];
-	let ary: [u8; 32] = U256::from(1).into();
+	let ary: [u8; 32] = U256::from(1).to_big_endian();
 	assert_eq!(ary, expected);
 }
 
 #[test]
 fn test_u256_from_fixed_array() {
 	let ary = [0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 123];
-	let num: U256 = ary.into();
+	let num = U256::from_big_endian(&ary);
 	assert_eq!(num, U256::from(core::u64::MAX) + 1 + 123);
 
-	let a_ref: &U256 = &ary.into();
+	let a_ref = &U256::from_big_endian(&ary);
 	assert_eq!(a_ref, &(U256::from(core::u64::MAX) + 1 + 123));
 }
 
@@ -1108,7 +1107,7 @@ fn test_u256_from_fixed_array() {
 fn test_from_ref_to_fixed_array() {
 	let ary: &[u8; 32] =
 		&[1, 0, 1, 2, 1, 0, 1, 2, 3, 0, 3, 4, 3, 0, 3, 4, 5, 0, 5, 6, 5, 0, 5, 6, 7, 0, 7, 8, 7, 0, 7, 8];
-	let big: U256 = ary.into();
+	let big = U256::from_big_endian(ary);
 	// the numbers are each row of 8 bytes reversed and cast to u64
 	assert_eq!(big, U256([504410889324070664, 360293493601469702, 216176097878868740, 72058702156267778u64]));
 }
@@ -1119,11 +1118,11 @@ fn test_u512_from_fixed_array() {
 		0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
 		0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 123,
 	];
-	let num: U512 = ary.into();
+	let num = U512::from_big_endian(&ary);
 	assert_eq!(num, U512::from(123));
 
-	let a_ref: &U512 = &ary.into();
-	assert_eq!(a_ref, &U512::from(123));
+	let a_ref = U512::from_big_endian(&ary);
+	assert_eq!(a_ref, U512::from(123));
 }
 
 #[test]
@@ -1138,7 +1137,7 @@ fn leading_zeros() {
 fn issue_507_roundtrip() {
 	let mut b32 = <[u8; 32]>::default();
 	let a = U256::from(10);
-	a.to_little_endian(&mut b32);
+	a.write_as_little_endian(&mut b32);
 	let b = U256::from_little_endian(&b32[..]);
 	assert_eq!(a, b);
 }