diff --git a/client_test.go b/client_test.go
index c77e0d93..e9c9a54e 100644
--- a/client_test.go
+++ b/client_test.go
@@ -4001,6 +4001,7 @@ func Test_Client_Maintenance(t *testing.T) {
 		t.Parallel()
 
 		config := newTestConfig(t, "")
+		config.RetryPolicy = &retrypolicytest.RetryPolicySlow{} // make sure jobs aren't worked before we can assert on job
 		config.RescueStuckJobsAfter = 5 * time.Minute
 
 		client, bundle := setup(t, config)
@@ -4036,11 +4037,12 @@ func Test_Client_Maintenance(t *testing.T) {
 		svc.TestSignals.FetchedBatch.WaitOrTimeout()
 		svc.TestSignals.UpdatedBatch.WaitOrTimeout()
 
-		requireJobHasState := func(jobID int64, state rivertype.JobState) {
+		requireJobHasState := func(jobID int64, state rivertype.JobState) *rivertype.JobRow {
 			t.Helper()
 			job, err := bundle.exec.JobGetByID(ctx, &riverdriver.JobGetByIDParams{ID: jobID, Schema: bundle.schema})
 			require.NoError(t, err)
 			require.Equal(t, state, job.State)
+			return job
 		}
 
 		// unchanged
@@ -4049,8 +4051,10 @@ func Test_Client_Maintenance(t *testing.T) {
 		requireJobHasState(ineligibleJob3.ID, ineligibleJob3.State)
 
 		// Jobs to retry should be retryable:
-		requireJobHasState(jobStuckToRetry1.ID, rivertype.JobStateRetryable)
-		requireJobHasState(jobStuckToRetry2.ID, rivertype.JobStateRetryable)
+		updatedJobStuckToRetry1 := requireJobHasState(jobStuckToRetry1.ID, rivertype.JobStateRetryable)
+		require.Greater(t, updatedJobStuckToRetry1.ScheduledAt, now.Add(10*time.Minute)) // make sure `scheduled_at` is a good margin in the future so it's not at risk of immediate retry (which could cause intermittent test issues)
+		updatedJobStuckToRetry2 := requireJobHasState(jobStuckToRetry2.ID, rivertype.JobStateRetryable)
+		require.Greater(t, updatedJobStuckToRetry2.ScheduledAt, now.Add(10*time.Minute))
 
 		// This one should be discarded because it's already at MaxAttempts:
 		requireJobHasState(jobStuckToDiscard.ID, rivertype.JobStateDiscarded)
diff --git a/internal/riverinternaltest/retrypolicytest/retrypolicytest.go b/internal/riverinternaltest/retrypolicytest/retrypolicytest.go
index 7b761223..05598543 100644
--- a/internal/riverinternaltest/retrypolicytest/retrypolicytest.go
+++ b/internal/riverinternaltest/retrypolicytest/retrypolicytest.go
@@ -59,7 +59,10 @@ func (p *RetryPolicyNoJitter) retrySecondsWithoutJitter(attempt int) float64 {
 	return min(retrySeconds, maxDurationSeconds)
 }
 
-const retryPolicySlowInterval = 1 * time.Hour
+// Choose a number here that's larger than the job rescuer's rescue interval of
+// one hour so that can easily use it to test the job rescuer (in addition to
+// other things).
+const retryPolicySlowInterval = 2 * time.Hour
 
 // RetryPolicySlow is a retry policy that has a very slow retry interval. This
 // is used in tests that check retries to make sure that in slower environments
diff --git a/internal/riverinternaltest/retrypolicytest/retrypolicytest_test.go b/internal/riverinternaltest/retrypolicytest/retrypolicytest_test.go
new file mode 100644
index 00000000..768513bc
--- /dev/null
+++ b/internal/riverinternaltest/retrypolicytest/retrypolicytest_test.go
@@ -0,0 +1,17 @@
+package retrypolicytest
+
+import (
+	"testing"
+	"time"
+
+	"github.com/stretchr/testify/require"
+)
+
+func TestRetryPolicySlowIntervalSufficientlyLarge(t *testing.T) {
+	t.Parallel()
+
+	// Should be a good margin larger than the job rescuer's default interval so
+	// that it can easily be used to test the job rescuer with making jobs
+	// accidentally eligible to work again right away.
+	require.Greater(t, (&RetryPolicySlow{}).Interval(), 1*time.Hour)
+}