diff --git a/tests/snippets/ints.py b/tests/snippets/ints.py
index 3db203a3e..baa145bff 100644
--- a/tests/snippets/ints.py
+++ b/tests/snippets/ints.py
@@ -49,6 +49,8 @@ with assert_raises(ZeroDivisionError):
 assert (-3).__rdivmod__(2) == (-1, -1)
 assert (2).__pow__(3) == 8
 assert (10).__pow__(-1) == 0.1
+with assert_raises(ZeroDivisionError):
+    (0).__pow__(-1)
 assert (2).__rpow__(3) == 9
 assert (10).__mod__(5) == 0
 assert (10).__mod__(6) == 4
diff --git a/vm/src/obj/objfloat.rs b/vm/src/obj/objfloat.rs
index 16b784847..fbcc6fcc0 100644
--- a/vm/src/obj/objfloat.rs
+++ b/vm/src/obj/objfloat.rs
@@ -154,6 +154,15 @@ fn inner_gt_int(value: f64, other_int: &BigInt) -> bool {
     }
 }
 
+pub fn float_pow(v1: f64, v2: f64, vm: &VirtualMachine) -> PyResult {
+    if v1.is_zero() {
+        let msg = format!("{} cannot be raised to a negative power", v1);
+        Err(vm.new_zero_division_error(msg))
+    } else {
+        v1.powf(v2).into_pyobject(vm)
+    }
+}
+
 #[pyimpl]
 #[allow(clippy::trivially_copy_pass_by_ref)]
 impl PyFloat {
@@ -359,7 +368,7 @@ impl PyFloat {
     fn pow(&self, other: PyObjectRef, vm: &VirtualMachine) -> PyResult {
         try_float(&other, vm)?.map_or_else(
             || Ok(vm.ctx.not_implemented()),
-            |other| self.value.powf(other).into_pyobject(vm),
+            |other| float_pow(self.value, other, vm),
         )
     }
 
@@ -367,7 +376,7 @@ impl PyFloat {
     fn rpow(&self, other: PyObjectRef, vm: &VirtualMachine) -> PyResult {
         try_float(&other, vm)?.map_or_else(
             || Ok(vm.ctx.not_implemented()),
-            |other| other.powf(self.value).into_pyobject(vm),
+            |other| float_pow(other, self.value, vm),
         )
     }
 
diff --git a/vm/src/obj/objint.rs b/vm/src/obj/objint.rs
index 43506b58d..61be01457 100644
--- a/vm/src/obj/objint.rs
+++ b/vm/src/obj/objint.rs
@@ -18,6 +18,7 @@ use crate::vm::VirtualMachine;
 use super::objbool::IntoPyBool;
 use super::objbyteinner::PyByteInner;
 use super::objbytes::PyBytes;
+use super::objfloat;
 use super::objint;
 use super::objstr::{PyString, PyStringRef};
 use super::objtype;
@@ -118,12 +119,12 @@ impl_try_from_object_int!(
 
 #[allow(clippy::collapsible_if)]
 fn inner_pow(int1: &PyInt, int2: &PyInt, vm: &VirtualMachine) -> PyResult {
-    let result = if int2.value.is_negative() {
+    if int2.value.is_negative() {
         let v1 = int1.float(vm)?;
         let v2 = int2.float(vm)?;
-        vm.ctx.new_float(v1.pow(v2))
+        objfloat::float_pow(v1, v2, vm)
     } else {
-        if let Some(v2) = int2.value.to_u64() {
+        Ok(if let Some(v2) = int2.value.to_u64() {
             vm.ctx.new_int(int1.value.pow(v2))
         } else if int1.value.is_one() || int1.value.is_zero() {
             vm.ctx.new_int(int1.value.clone())
@@ -137,9 +138,8 @@ fn inner_pow(int1: &PyInt, int2: &PyInt, vm: &VirtualMachine) -> PyResult {
             // missing feature: BigInt exp
             // practically, exp over u64 is not possible to calculate anyway
             vm.ctx.not_implemented()
-        }
-    };
-    Ok(result)
+        })
+    }
 }
 
 fn inner_mod(int1: &PyInt, int2: &PyInt, vm: &VirtualMachine) -> PyResult {