diff --git a/docs/CHANGELOG.md b/docs/CHANGELOG.md
index 1aa1d1213f..8fba25b80c 100644
--- a/docs/CHANGELOG.md
+++ b/docs/CHANGELOG.md
@@ -3,6 +3,7 @@
 # tip
 
 * BUGFIX: vmagent: prevent from `dialing to the given TCP address time out` error when scraping big number of unavailable targets. See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/987
+* BUGFIX: vmagent: properly show scrape duration on `/targets` page. Previously it was incorrectly shown as 0.000s.
 * BUGFIX: vmalert: return non-empty result in template func `query` stub to pass validation. See https://github.com/VictoriaMetrics/VictoriaMetrics/issues/989 .
 
 * FEATURE: provide a sample list of alerting rules for VictoriaMetrics components. It is available [here](https://github.com/VictoriaMetrics/VictoriaMetrics/blob/master/deployment/docker/alerts.yml).
diff --git a/lib/promscrape/targetstatus.go b/lib/promscrape/targetstatus.go
index f8804668a6..47b26f4df3 100644
--- a/lib/promscrape/targetstatus.go
+++ b/lib/promscrape/targetstatus.go
@@ -312,7 +312,7 @@ func (tsm *targetStatusMap) getTargetsStatusByJob() []jobTargetsStatuses {
 				labels:         promrelabel.FinalizeLabels(nil, st.sw.Labels),
 				originalLabels: st.sw.OriginalLabels,
 				lastScrapeTime: st.getDurationFromLastScrape(),
-				scrapeDuration: time.Duration(st.scrapeDuration),
+				scrapeDuration: time.Duration(st.scrapeDuration) * time.Millisecond,
 				error:          errMsg,
 			})
 		}