diff --git a/libcxx/include/random b/libcxx/include/random
--- a/libcxx/include/random
+++ b/libcxx/include/random
@@ -1668,7 +1668,23 @@
 
 template <unsigned long long __a, unsigned long long __c,
           unsigned long long __m, unsigned long long _Mp,
-          bool _MightOverflow = (__a != 0 && __m != 0 && __m-1 > (_Mp-__c)/__a)>
+          bool _MightOverflow = (__a != 0 && __m != 0 && __m-1 > (_Mp-__c)/__a),
+          bool _OverflowOK = ((__m|__m-1) > __m), // m = 2^n
+          bool _SchrageOK = (__a != 0 && __m != 0 && __m % __a <= __m / __a)> // r <= q
+struct __lce_alg_picker
+{
+    static_assert(__a != 0 || __m != 0 || !_MightOverflow || _OverflowOK || _SchrageOK,
+                  "The current values of a, c, and m cannot generate a number "
+                  "within bounds of linear_congruential_engine.");
+
+    static _LIBCPP_CONSTEXPR const bool __use_schrage = _MightOverflow &&
+                                                        !_OverflowOK &&
+                                                        _SchrageOK;
+};
+
+template <unsigned long long __a, unsigned long long __c,
+          unsigned long long __m, unsigned long long _Mp,
+          bool _UseSchrage = __lce_alg_picker<__a, __c, __m, _Mp>::__use_schrage>
 struct __lce_ta;
 
 // 64
@@ -1842,6 +1858,7 @@
 
     static_assert(__m == 0 || __a < __m, "linear_congruential_engine invalid parameters");
     static_assert(__m == 0 || __c < __m, "linear_congruential_engine invalid parameters");
+    static_assert(_VSTD::is_unsigned<_UIntType>::value, "_UIntType must be unsigned type");
 public:
     static _LIBCPP_CONSTEXPR const result_type _Min = __c == 0u ? 1u: 0u;
     static _LIBCPP_CONSTEXPR const result_type _Max = __m - 1u;
diff --git a/libcxx/test/std/numerics/rand/rand.eng/rand.eng.lcong/alg.pass.cpp b/libcxx/test/std/numerics/rand/rand.eng/rand.eng.lcong/alg.pass.cpp
new file mode 100644
--- /dev/null
+++ b/libcxx/test/std/numerics/rand/rand.eng/rand.eng.lcong/alg.pass.cpp
@@ -0,0 +1,69 @@
+//===----------------------------------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+// <random>
+
+// template <class UIntType, UIntType a, UIntType c, UIntType m>
+//   class linear_congruential_engine;
+
+// result_type operator()();
+
+#include <random>
+#include <cassert>
+
+#include "test_macros.h"
+
+int main(int, char**)
+{
+    typedef unsigned long long T;
+
+    // m might overflow, but the overflow is OK so it shouldn't use schrage's algorithm
+    typedef std::linear_congruential_engine<T, 25214903917ull, 1, (1ull<<48)> E1;
+    E1 e1;
+    // make sure the right algorithm was used
+    assert(e1() == 25214903918);
+    assert(e1() == 205774354444503);
+    assert(e1() == 158051849450892);
+    // make sure result is in bounds
+    assert(e1() < (1ull<<48));
+    assert(e1() < (1ull<<48));
+    assert(e1() < (1ull<<48));
+    assert(e1() < (1ull<<48));
+    assert(e1() < (1ull<<48));
+
+    // m might overflow. The overflow is not OK and result will be in bounds
+    // so we should use shrage's algorithm
+    typedef std::linear_congruential_engine<T, (1ull<<2), 0, (1ull<<63) + 1> E2;
+    E2 e2;
+    // make sure shrage's algorithm is used (it would be 0s otherwise)
+    assert(e2() == 4);
+    assert(e2() == 16);
+    assert(e2() == 64);
+    // make sure result is in bounds
+    assert(e2() < (1ull<<48) + 1);
+    assert(e2() < (1ull<<48) + 1);
+    assert(e2() < (1ull<<48) + 1);
+    assert(e2() < (1ull<<48) + 1);
+    assert(e2() < (1ull<<48) + 1);
+
+    // m will not overflow so we should not use shrage's algorithm
+    typedef std::linear_congruential_engine<T, 1ull, 1, (1ull<<48)> E3;
+    E3 e3;
+    // make sure the correct algorithm was used
+    assert(e3() == 2);
+    assert(e3() == 3);
+    assert(e3() == 4);
+    // make sure result is in bounds
+    assert(e3() < (1ull<<48));
+    assert(e3() < (1ull<<48));
+    assert(e3() < (1ull<<48));
+    assert(e3() < (1ull<<48));
+    assert(e2() < (1ull<<48));
+
+  return 0;
+}
\ No newline at end of file
diff --git a/libcxx/test/std/numerics/rand/rand.eng/rand.eng.lcong/params.fail.cpp b/libcxx/test/std/numerics/rand/rand.eng/rand.eng.lcong/params.fail.cpp
new file mode 100644
--- /dev/null
+++ b/libcxx/test/std/numerics/rand/rand.eng/rand.eng.lcong/params.fail.cpp
@@ -0,0 +1,34 @@
+//===----------------------------------------------------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+// <random>
+
+// template <class UIntType, UIntType a, UIntType c, UIntType m>
+//   class linear_congruential_engine;
+
+// requirements on parameters
+
+#include <random>
+#include <cassert>
+
+#include "test_macros.h"
+
+int main(int, char**)
+{
+    typedef unsigned long long T;
+
+    // expected-error@random:* {{static_assert failed due to requirement '1ULL == 0 || 1ULL < 1ULL' "linear_congruential_engine invalid parameters"}}
+    std::linear_congruential_engine<T, 0, 0, 0> e2;
+    // expected-error@random:* {{static_assert failed due to requirement '1ULL == 0 || 1ULL < 1ULL' "linear_congruential_engine invalid parameters"}}
+    std::linear_congruential_engine<T, 0, 1, 1> e3;
+    std::linear_congruential_engine<T, 1, 0, 1> e4;
+    // expected-error@random:* {{static_assert failed due to requirement 'std::__1::is_unsigned<int>::value' "_UIntType must be unsigned type"}}
+    std::linear_congruential_engine<int, 0, 0, 0> e5;
+
+  return 0;
+}
\ No newline at end of file