diff --git a/docs/gitbook/dev/upgrade-guide.md b/docs/gitbook/dev/upgrade-guide.md
index 0c5aeff2c..157fbe233 100644
--- a/docs/gitbook/dev/upgrade-guide.md
+++ b/docs/gitbook/dev/upgrade-guide.md
@@ -49,7 +49,7 @@ Istio 1.5 comes with a breaking change for Flagger uses. In Istio telemetry v2 t
 `istio_request_duration_seconds_bucket` has been removed and replaced with `istio_request_duration_milliseconds_bucket`
 and this breaks the `request-duration` metric check.
 
-You can create a metric template using the new duration metric like this:
+If are using **Istio 1.4**, you can create a metric template using the old duration metric like this:
 
 ```yaml
 apiVersion: flagger.app/v1beta1
@@ -66,7 +66,7 @@ spec:
         0.99,
         sum(
             rate(
-                istio_request_duration_milliseconds_bucket{
+                istio_request_duration_seconds_bucket{
                     reporter="destination",
                     destination_workload_namespace="{{ namespace }}",
                     destination_workload=~"{{ target }}"
@@ -85,6 +85,6 @@ metrics:
     name: latency
     namespace: istio-system
   thresholdRange:
-    max: 500
+    max: 0.500
   interval: 1m
 ```
diff --git a/docs/gitbook/tutorials/istio-progressive-delivery.md b/docs/gitbook/tutorials/istio-progressive-delivery.md
index 39b6df63e..26763fb83 100644
--- a/docs/gitbook/tutorials/istio-progressive-delivery.md
+++ b/docs/gitbook/tutorials/istio-progressive-delivery.md
@@ -6,7 +6,7 @@ This guide shows you how to use Istio and Flagger to automate canary deployments
 
 ## Prerequisites
 
-Flagger requires a Kubernetes cluster **v1.11** or newer and Istio **v1.0** or newer.
+Flagger requires a Kubernetes cluster **v1.11** or newer and Istio **v1.5** or newer.
 
 Install Istio with telemetry support and Prometheus:
 
@@ -14,7 +14,7 @@ Install Istio with telemetry support and Prometheus:
 istioctl manifest apply --set profile=default
 ```
 
-Install Flagger using Kustomize (kubectl 1.14) in the `istio-system` namespace:
+Install Flagger using Kustomize (kubectl >= 1.14) in the `istio-system` namespace:
 
 ```bash
 kubectl apply -k github.com/weaveworks/flagger//kustomize/istio
@@ -149,7 +149,7 @@ spec:
           cmd: "hey -z 1m -q 10 -c 2 http://podinfo-canary.test:9898/"
 ```
 
-**Note** that when using Istio 1.5 you have to replace the `request-duration`
+**Note** that when using Istio 1.4 you have to replace the `request-duration`
 with a [metric template](https://docs.flagger.app/dev/upgrade-guide#istio-telemetry-v2).
 
 Save the above resource as podinfo-canary.yaml and then apply it:
diff --git a/pkg/metrics/observers/istio.go b/pkg/metrics/observers/istio.go
index 26084451f..baf4084f7 100644
--- a/pkg/metrics/observers/istio.go
+++ b/pkg/metrics/observers/istio.go
@@ -36,7 +36,7 @@ var istioQueries = map[string]string{
 		0.99,
 		sum(
 			rate(
-				istio_request_duration_seconds_bucket{
+				istio_request_duration_milliseconds_bucket{
 					reporter="destination",
 					destination_workload_namespace="{{ namespace }}",
 					destination_workload=~"{{ target }}"
@@ -75,6 +75,6 @@ func (ob *IstioObserver) GetRequestDuration(model flaggerv1.MetricTemplateModel)
 		return 0, fmt.Errorf("running query failed: %w", err)
 	}
 
-	ms := time.Duration(int64(value*1000)) * time.Millisecond
+	ms := time.Duration(int64(value)) * time.Millisecond
 	return ms, nil
 }
diff --git a/pkg/metrics/observers/istio_test.go b/pkg/metrics/observers/istio_test.go
index 2bd829d6e..eab6b8bee 100644
--- a/pkg/metrics/observers/istio_test.go
+++ b/pkg/metrics/observers/istio_test.go
@@ -49,13 +49,13 @@ func TestIstioObserver_GetRequestSuccessRate(t *testing.T) {
 }
 
 func TestIstioObserver_GetRequestDuration(t *testing.T) {
-	expected := ` histogram_quantile( 0.99, sum( rate( istio_request_duration_seconds_bucket{ reporter="destination", destination_workload_namespace="default", destination_workload=~"podinfo" }[1m] ) ) by (le) )`
+	expected := ` histogram_quantile( 0.99, sum( rate( istio_request_duration_milliseconds_bucket{ reporter="destination", destination_workload_namespace="default", destination_workload=~"podinfo" }[1m] ) ) by (le) )`
 
 	ts := httptest.NewServer(http.HandlerFunc(func(w http.ResponseWriter, r *http.Request) {
 		promql := r.URL.Query()["query"][0]
 		assert.Equal(t, expected, promql)
 
-		json := `{"status":"success","data":{"resultType":"vector","result":[{"metric":{},"value":[1,"0.100"]}]}}`
+		json := `{"status":"success","data":{"resultType":"vector","result":[{"metric":{},"value":[1,"100"]}]}}`
 		w.Write([]byte(json))
 	}))
 	defer ts.Close()