diff --git a/noir_stdlib/src/lib.nr b/noir_stdlib/src/lib.nr
index 9b166f6ae94..1430cb5788e 100644
--- a/noir_stdlib/src/lib.nr
+++ b/noir_stdlib/src/lib.nr
@@ -23,6 +23,7 @@ mod cmp;
 mod ops;
 mod default;
 mod prelude;
+mod uint128;
 
 // Oracle calls are required to be wrapped in an unconstrained function
 // Thus, the only argument to the `println` oracle is expected to always be an ident
diff --git a/noir_stdlib/src/prelude.nr b/noir_stdlib/src/prelude.nr
index 56020509122..b57ff460371 100644
--- a/noir_stdlib/src/prelude.nr
+++ b/noir_stdlib/src/prelude.nr
@@ -1,5 +1,6 @@
 use crate::collections::vec::Vec;
 use crate::option::Option;
 use crate::{print, println, assert_constant};
+use crate::uint128::U128;
 use crate::cmp::{Eq, Ord};
 use crate::default::Default;
diff --git a/noir_stdlib/src/uint128.nr b/noir_stdlib/src/uint128.nr
new file mode 100644
index 00000000000..4a58b3868be
--- /dev/null
+++ b/noir_stdlib/src/uint128.nr
@@ -0,0 +1,292 @@
+use crate::ops::{Add, Sub, Mul, Div, Rem, BitOr, BitAnd, BitXor, Shl, Shr};
+use crate::cmp::{Eq, Ord, Ordering};
+
+global pow64 : Field = 18446744073709551616; //2^64;
+
+struct U128 {
+    lo: Field,
+    hi: Field,
+}
+
+impl U128 {
+
+    pub fn from_u64s_le(lo: u64, hi: u64) -> U128 {
+        // in order to handle multiplication, we need to represent the product of two u64 without overflow
+        assert(crate::field::modulus_num_bits() as u32 > 128);
+        U128 {
+            lo: lo as Field,
+            hi: hi as Field,
+        }
+    }
+
+    pub fn from_u64s_be(hi: u64, lo: u64) -> U128 {
+        U128::from_u64s_le(lo,hi)
+    }
+
+    pub fn from_le_bytes(bytes: [u8; 16]) -> U128 {
+        let mut lo = 0;
+        let mut base = 1;
+        for i in 0..8 {
+            lo += (bytes[i] as Field)*base;
+            base *= 256;
+        }
+        let mut hi = 0;
+        base = 1;
+        for i in 8..16 {
+            hi += (bytes[i] as Field)*base;
+            base *= 256;
+        }
+        U128 {
+            lo,
+            hi,
+        }
+    }
+
+    pub fn to_le_bytes(self: Self) -> [u8; 16] {
+        let lo = self.lo.to_le_bytes(8);
+        let hi = self.hi.to_le_bytes(8);
+        let mut bytes = [0;16];
+        for i in 0..8 {
+            bytes[i] = lo[i];
+            bytes[i+8] = hi[i];
+        }
+        bytes
+    }
+
+    pub fn from_hex<N>(hex: str<N>) -> U128 {
+        let N = N as u32;
+        let bytes = hex.as_bytes();
+        // string must starts with "0x"
+        assert((bytes[0] == 48) & (bytes[1] == 120), "Invalid hexadecimal string");
+        assert(N < 35, "Input does not fit into a U128");
+
+        let mut lo = 0;
+        let mut hi = 0;
+        let mut base = 1; 
+        if N <= 18 {
+            for i in 0..N-2 {
+                lo += U128::decode_ascii(bytes[N-i-1])*base;
+                base = base*16;
+            }
+        } else {
+            for i in 0..16 {
+                lo += U128::decode_ascii(bytes[N-i-1])*base;
+                base = base*16;
+            }
+            base = 1;
+            for i in 17..N-1 {           
+                hi += U128::decode_ascii(bytes[N-i])*base;
+                base = base*16;
+            }
+        }
+        U128 {
+            lo: lo as Field,
+            hi: hi as Field,
+        }
+    }
+
+    fn decode_ascii(ascii: u8) -> Field {
+        if ascii < 58 {
+            ascii - 48
+        } else {
+            if ascii < 71 {
+                ascii - 55
+            } else {
+                ascii - 87
+            }
+            
+        } as Field
+    }
+
+    unconstrained fn unconstrained_div(self: Self, b: U128) -> (U128, U128) {
+        if self < b {
+            (U128::from_u64s_le(0, 0), self)
+        } else {
+            //TODO check if this can overflow?
+            let (q,r) = self.unconstrained_div(b * U128::from_u64s_le(2,0));
+            let q_mul_2 = q * U128::from_u64s_le(2,0);
+            if r < b {
+                (q_mul_2, r)
+            } else {
+                (q_mul_2 + U128::from_u64s_le(1,0), r - b)
+            }
+
+        }  
+    }
+
+    pub fn from_integer<T>(i: T) -> U128 {
+        let f = crate::as_field(i);
+        let lo = f as u64 as Field;
+        let hi = (f-lo) / pow64;
+        U128 {
+            lo,
+            hi,
+        }
+    }
+
+    pub fn to_integer<T>(self) -> T {
+        crate::from_field(self.lo+self.hi*pow64)
+    }
+
+    fn wrapping_mul(self: Self, b: U128) -> U128 {
+        let low = self.lo*b.lo;
+        let lo = low as u64 as Field;
+        let carry = (low - lo) / pow64;
+        let high = if crate::field::modulus_num_bits() as u32 > 196 {
+            (self.lo+self.hi)*(b.lo+b.hi) - low + carry
+        } else {
+            self.lo*b.hi + self.hi*b.lo + carry
+        };
+        let hi = high as u64 as Field;
+        U128 {
+            lo,
+            hi,
+        }
+    }
+}
+
+impl Add for U128 {
+    pub fn add(self: Self, b: U128) -> U128 {
+        let low = self.lo + b.lo;
+        let lo = low as u64 as Field;
+        let carry = (low - lo) / pow64;  
+        let high = self.hi + b.hi + carry;
+        let hi = high as u64 as Field;
+        assert(hi == high, "attempt to add with overflow");
+        U128 {
+            lo,
+            hi,
+        }
+    }
+}
+
+impl Sub for U128 {
+    pub fn sub(self: Self, b: U128) -> U128 {
+        let low = pow64 + self.lo - b.lo;
+        let lo = low as u64 as Field;
+        let borrow = (low == lo) as Field;
+        let high = self.hi - b.hi - borrow;
+        let hi = high as u64 as Field;
+        assert(hi == high, "attempt to subtract with overflow");
+        U128 {
+            lo,
+            hi,
+        }
+    }
+}
+
+impl Mul for U128 {
+    pub fn mul(self: Self, b: U128) -> U128 {
+        assert(self.hi*b.hi == 0, "attempt to multiply with overflow");
+        let low = self.lo*b.lo;
+        let lo = low as u64 as Field;
+        let carry = (low - lo) / pow64;
+        let high = if crate::field::modulus_num_bits() as u32 > 196 {
+            (self.lo+self.hi)*(b.lo+b.hi) - low + carry
+        } else {
+            self.lo*b.hi + self.hi*b.lo + carry
+        };
+        let hi = high as u64 as Field;
+        assert(hi == high, "attempt to multiply with overflow");
+        U128 {
+            lo,
+            hi,
+        }
+    }
+}
+
+impl Div for U128 {
+    pub fn div(self: Self, b: U128) -> U128 {
+        let (q,r) = self.unconstrained_div(b);
+        let a = b * q + r;
+        assert_eq(self, a);
+        assert(r < b);
+        q
+    }
+}
+
+impl Rem for U128 {
+    pub fn rem(self: Self, b: U128) -> U128 {
+        let (q,r) = self.unconstrained_div(b);
+        let a = b * q + r;
+        assert_eq(self, a);
+        assert(r < b);
+        r
+    }
+}
+
+impl Eq for U128 {
+    pub fn eq(self: Self, b: U128) -> bool {
+        (self.lo == b.lo) & (self.hi == b.hi)
+    }
+}
+
+impl Ord for U128 {
+    fn cmp(self, other: Self) -> Ordering {
+        let hi_ordering = (self.hi as u64).cmp((other.hi as u64));
+        let lo_ordering = (self.lo as u64).cmp((other.lo as u64));
+        
+        if hi_ordering == Ordering::equal() {
+            lo_ordering
+        } else {
+            hi_ordering
+        }
+    }
+}
+
+impl BitOr for U128 { 
+    fn bitor(self, other: U128) -> U128 {
+        U128 {
+            lo: ((self.lo as u64) | (other.lo as u64)) as Field,
+            hi: ((self.hi as u64) | (other.hi as u64))as Field
+        }
+    }
+}
+
+impl BitAnd for U128 {
+    fn bitand(self, other: U128) -> U128 { 
+        U128 {
+            lo: ((self.lo as u64) & (other.lo as u64)) as Field,
+            hi: ((self.hi as u64) & (other.hi as u64)) as Field
+        }
+    }
+}
+
+impl BitXor for U128 {
+    fn bitxor(self, other: U128) -> U128 { 
+        U128 {
+            lo: ((self.lo as u64) ^ (other.lo as u64)) as Field,
+            hi: ((self.hi as u64) ^ (other.hi as u64)) as Field
+        }
+    }
+}
+
+impl Shl for U128 { 
+    fn shl(self, other: U128) -> U128 { 
+        assert(other < U128::from_u64s_le(128,0), "attempt to shift left with overflow");
+        let exp_bits = other.lo.to_be_bits(7);
+
+        let mut r: Field = 2;
+        let mut y: Field = 1;
+        for i in 1..8 {
+            y = (exp_bits[7-i] as Field) * (r * y) + (1 - exp_bits[7-i] as Field) * y;
+            r *= r;
+        }
+        self.wrapping_mul(U128::from_integer(y))
+    } 
+}
+
+impl Shr for U128 { 
+    fn shr(self, other: U128) -> U128 { 
+        assert(other < U128::from_u64s_le(128,0), "attempt to shift right with overflow");
+        let exp_bits = other.lo.to_be_bits(7);
+
+        let mut r: Field = 2;
+        let mut y: Field = 1;
+        for i in 1..8 {
+            y = (exp_bits[7-i] as Field) * (r * y) + (1 - exp_bits[7-i] as Field) * y;
+            r *= r;
+        }
+        self / U128::from_integer(y)
+    } 
+}
\ No newline at end of file
diff --git a/test_programs/execution_success/u128/Nargo.toml b/test_programs/execution_success/u128/Nargo.toml
new file mode 100644
index 00000000000..c1dcd84db04
--- /dev/null
+++ b/test_programs/execution_success/u128/Nargo.toml
@@ -0,0 +1,6 @@
+[package]
+name = "u128"
+type = "bin"
+authors = [""]
+
+[dependencies]
diff --git a/test_programs/execution_success/u128/Prover.toml b/test_programs/execution_success/u128/Prover.toml
new file mode 100644
index 00000000000..961db9825a7
--- /dev/null
+++ b/test_programs/execution_success/u128/Prover.toml
@@ -0,0 +1,7 @@
+x = "3"
+y = "4"
+z = "7"
+hexa ="0x1f03a"
+[big_int]
+lo = 1
+hi = 2
\ No newline at end of file
diff --git a/test_programs/execution_success/u128/src/main.nr b/test_programs/execution_success/u128/src/main.nr
new file mode 100644
index 00000000000..4c734f3a8f9
--- /dev/null
+++ b/test_programs/execution_success/u128/src/main.nr
@@ -0,0 +1,44 @@
+use dep::std;
+
+fn main(mut x: u32, y: u32, z: u32, big_int: U128, hexa: str<7>) {
+    let a = U128::from_u64s_le(x as u64, x as u64);
+    let b = U128::from_u64s_le(y as u64, x as u64);
+    let c = a + b;
+    assert(c.lo == z as Field);
+    assert(c.hi == 2 * x as Field);
+    assert(U128::from_hex(hexa).lo == 0x1f03a);
+    let t1 = U128::from_hex("0x9d9c7a87771f03a23783f9d9c7a8777");
+    let t2 = U128::from_hex("0x45a26c708BFCF39041");
+    let t = t1 + t2;
+    assert(t.lo == 0xc5e4b029996e17b8);
+    assert(t.hi == 0x09d9c7a87771f07f);
+    let t3 = U128::from_le_bytes(t.to_le_bytes());
+    assert(t == t3);
+
+    let t4 = t - t2;
+    assert(t4 == t1);
+
+    let t5 = U128::from_u64s_le(0, 1);
+    let t6 = U128::from_u64s_le(1, 0);
+    assert((t5 - t6).hi == 0);
+
+    assert(
+        (U128::from_hex("0x71f03a23783f9d9c7a8777") * U128::from_hex("0x8BFCF39041")).hi
+        == U128::from_hex("0x3e4e0471b873470e247c824e61445537").hi
+    );
+    let q = U128::from_hex("0x3e4e0471b873470e247c824e61445537") / U128::from_hex("0x8BFCF39041");
+    assert(q == U128::from_hex("0x71f03a23783f9d9c7a8777"));
+
+    assert(big_int.hi == 2);
+
+    let mut small_int = U128::from_integer(x);
+    assert(small_int.lo == x as Field);
+    assert(x == small_int.to_integer());
+    let shift = small_int << small_int;
+    assert(shift == U128::from_integer(x << x));
+    assert(shift >> small_int == small_int);
+    assert(shift >> U128::from_integer(127) == U128::from_integer(0));
+    assert(shift << U128::from_integer(127) == U128::from_integer(0));
+
+}
+