diff --git a/cpp/src/gandiva/function_registry_math_ops.cc b/cpp/src/gandiva/function_registry_math_ops.cc
index 968d04bfbd925..f3d80110152d5 100644
--- a/cpp/src/gandiva/function_registry_math_ops.cc
+++ b/cpp/src/gandiva/function_registry_math_ops.cc
@@ -91,9 +91,6 @@ std::vector<NativeFunction> GetMathOpsFunctionRegistry() {
       UNARY_SAFE_NULL_IF_NULL(abs, {}, float32, float32),
       UNARY_SAFE_NULL_IF_NULL(abs, {}, float64, float64),
 
-      BINARY_GENERIC_SAFE_NULL_IF_NULL(round, {}, float64, int32, float64),
-      BINARY_GENERIC_SAFE_NULL_IF_NULL(round, {}, float64, int64, float64),
-
       // decimal functions
       UNARY_SAFE_NULL_IF_NULL(abs, {}, decimal128, decimal128),
       UNARY_SAFE_NULL_IF_NULL(ceil, {}, decimal128, decimal128),
diff --git a/cpp/src/gandiva/precompiled/extended_math_ops.cc b/cpp/src/gandiva/precompiled/extended_math_ops.cc
index 696fb4fd012e9..1ec80c195771a 100644
--- a/cpp/src/gandiva/precompiled/extended_math_ops.cc
+++ b/cpp/src/gandiva/precompiled/extended_math_ops.cc
@@ -26,7 +26,6 @@ extern "C" {
 
 #include <math.h>
 #include <stdio.h>
-#include <stddef.h>
 #include <stdlib.h>
 #include <string.h>
 
@@ -97,23 +96,6 @@ ABS_TYPES_UNARY(int64, uint64)
 ABS_FTYPES_UNARY(float32, float32)
 ABS_FTYPES_UNARY(float64, float64)
 
-// round
-#define ROUND_TYPES_UNARY(IN_TYPE1, IN_TYPE2, OUT_TYPE)                                 \
-  FORCE_INLINE                                                                          \
-  gdv_##OUT_TYPE round_##IN_TYPE1##_##IN_TYPE2(gdv_##IN_TYPE1 val, gdv_##IN_TYPE2 dp) { \
-    int charsNeeded = 1 + snprintf(NULL, 0, "%.*f", (int) dp, val);                     \
-    char* buffer = reinterpret_cast<char*>(malloc(charsNeeded));                        \
-    snprintf(buffer, charsNeeded, "%.*f", (int) dp, nextafter(val, val*2));             \
-    double result = atof(buffer);                                                       \
-    free(buffer);                                                                       \
-    return static_cast<gdv_##OUT_TYPE>(result);                                         \
-  }
-
-ROUND_TYPES_UNARY(float64, int32, float64)
-ROUND_TYPES_UNARY(float64, int64, float64)
-
-#undef ROUND_TYPES_UNARY
-
 FORCE_INLINE
 void set_error_for_logbase(int64_t execution_context, double base) {
   char const* prefix = "divide by zero error with log of base";
diff --git a/cpp/src/gandiva/precompiled/extended_math_ops_test.cc b/cpp/src/gandiva/precompiled/extended_math_ops_test.cc
index e8f24d9d139f5..3e9d8a5d2cd44 100644
--- a/cpp/src/gandiva/precompiled/extended_math_ops_test.cc
+++ b/cpp/src/gandiva/precompiled/extended_math_ops_test.cc
@@ -407,10 +407,4 @@ TEST(TestExtendedMathOps, TestBinRepresentation) {
             "1000000000000000000000000000000000000000000000000000000000000000");
   EXPECT_FALSE(ctx.has_error());
 }
-TEST(TestExtendedMathOps, TestRound) {
-  EXPECT_EQ(round_float64_int32(1234.56789, 4), 1234.5679);
-  EXPECT_EQ(round_float64_int64(1234.56789, 4), 1234.5679);
-  EXPECT_EQ(round_float64_int32(-1234.56789, 4), -1234.5679);
-  EXPECT_EQ(round_float64_int64(-1234.56789, 4), -1234.5679);
-}
 }  // namespace gandiva
diff --git a/cpp/src/gandiva/precompiled/types.h b/cpp/src/gandiva/precompiled/types.h
index e4e9ee0de313b..01c8cd6e67a7d 100644
--- a/cpp/src/gandiva/precompiled/types.h
+++ b/cpp/src/gandiva/precompiled/types.h
@@ -297,9 +297,6 @@ gdv_float64 degrees_int64(gdv_int64);
 gdv_float64 degrees_float32(gdv_float32);
 gdv_float64 degrees_float64(gdv_float64);
 
-gdv_float64 round_float64_int32(gdv_float64 in1, gdv_int32 in2);
-gdv_float64 round_float64_int64(gdv_float64 in1, gdv_int64 in2);
-
 gdv_int32 bitwise_and_int32_int32(gdv_int32 in1, gdv_int32 in2);
 gdv_int64 bitwise_and_int64_int64(gdv_int64 in1, gdv_int64 in2);
 gdv_int32 bitwise_or_int32_int32(gdv_int32 in1, gdv_int32 in2);