diff --git a/gossipsub.go b/gossipsub.go
index 68bef4a5..68bc5e18 100644
--- a/gossipsub.go
+++ b/gossipsub.go
@@ -74,6 +74,8 @@ var (
 	GossipSubIWantFollowupTime                = 3 * time.Second
 	GossipSubIDontWantMessageThreshold        = 1024 // 1KB
 	GossipSubIDontWantMessageTTL              = 3    // 3 heartbeats
+
+	GossipSubMaxIWantsPerMessageIDPerHeartbeat = 10
 )
 
 type checksum struct {
@@ -239,6 +241,11 @@ type GossipSubParams struct {
 
 	// IDONTWANT is cleared when it's older than the TTL.
 	IDontWantMessageTTL int
+
+	// MaxIWantsPerMessageIDPerHeartbeat is the maximum number of pending IWANT
+	// requests allowed per message ID per heartbeat. This helps limit the
+	// number of duplicates we'll receive from peers.
+	MaxIWantsPerMessageIDPerHeartbeat int
 }
 
 // NewGossipSub returns a new PubSub object using the default GossipSubRouter as the router.
@@ -276,6 +283,10 @@ func DefaultGossipSubRouter(h host.Host) *GossipSubRouter {
 		feature:      GossipSubDefaultFeatures,
 		tagTracer:    newTagTracer(h.ConnManager()),
 		params:       params,
+		// number of allowed IWANTs per message ID. If the message ID is
+		// missing, it must be initialized to
+		// `MaxIWantsPerMessageIDPerHeartbeat`
+		allowedIWantCount: make(map[string]int),
 	}
 }
 
@@ -315,6 +326,8 @@ func DefaultGossipSubParams() GossipSubParams {
 		IDontWantMessageThreshold: GossipSubIDontWantMessageThreshold,
 		IDontWantMessageTTL:       GossipSubIDontWantMessageTTL,
 		SlowHeartbeatWarning:      0.1,
+
+		MaxIWantsPerMessageIDPerHeartbeat: GossipSubMaxIWantsPerMessageIDPerHeartbeat,
 	}
 }
 
@@ -479,6 +492,10 @@ type GossipSubRouter struct {
 	connect      chan connectInfo                 // px connection requests
 	cab          peerstore.AddrBook
 
+	// allowed number of IWANTs per message ID. Must be initialized to
+	// `MaxIWantsPerMessageIDPerHeartbeat` if message id is missing.
+	allowedIWantCount map[string]int
+
 	protos  []protocol.ID
 	feature GossipSubFeatureTest
 
@@ -713,11 +730,13 @@ func (gs *GossipSubRouter) AcceptFrom(p peer.ID) AcceptStatus {
 func (gs *GossipSubRouter) Preprocess(from peer.ID, msgs []*Message) {
 	tmids := make(map[string][]string)
 	for _, msg := range msgs {
+		mid := gs.p.idGen.ID(msg)
+		delete(gs.allowedIWantCount, mid)
 		if len(msg.GetData()) < gs.params.IDontWantMessageThreshold {
 			continue
 		}
 		topic := msg.GetTopic()
-		tmids[topic] = append(tmids[topic], gs.p.idGen.ID(msg))
+		tmids[topic] = append(tmids[topic], mid)
 	}
 	for topic, mids := range tmids {
 		if len(mids) == 0 {
@@ -803,7 +822,20 @@ func (gs *GossipSubRouter) handleIHave(p peer.ID, ctl *pb.ControlMessage) []*pb.
 			if gs.p.seenMessage(mid) {
 				continue
 			}
+
+			allowedIWants, ok := gs.allowedIWantCount[mid]
+			if !ok {
+				allowedIWants = gs.params.MaxIWantsPerMessageIDPerHeartbeat
+			}
+
+			// Check if we've exceeded the maximum number of pending IWANTs for this message
+			if allowedIWants <= 0 {
+				log.Debugf("IHAVE: ignoring IHAVE for message %s from peer %s; too many inflight IWANTs", mid, p)
+				continue
+			}
+
 			iwant[mid] = struct{}{}
+			gs.allowedIWantCount[mid] = allowedIWants - 1
 		}
 	}
 
@@ -1453,6 +1485,9 @@ func (gs *GossipSubRouter) heartbeat() {
 	toprune := make(map[peer.ID][]string)
 	noPX := make(map[peer.ID]bool)
 
+	// reset number of allowed IWANT requests
+	gs.resetAllowedIWants()
+
 	// clean up expired backoffs
 	gs.clearBackoff()
 
@@ -1702,6 +1737,13 @@ func (gs *GossipSubRouter) heartbeat() {
 	gs.mcache.Shift()
 }
 
+func (gs *GossipSubRouter) resetAllowedIWants() {
+	if len(gs.allowedIWantCount) > 0 {
+		// throw away the old map and make a new one
+		gs.allowedIWantCount = make(map[string]int)
+	}
+}
+
 func (gs *GossipSubRouter) clearIHaveCounters() {
 	if len(gs.peerhave) > 0 {
 		// throw away the old map and make a new one
diff --git a/gossipsub_test.go b/gossipsub_test.go
index 9f450d87..ca87e454 100644
--- a/gossipsub_test.go
+++ b/gossipsub_test.go
@@ -3860,3 +3860,94 @@ func BenchmarkSplitRPCLargeMessages(b *testing.B) {
 		}
 	})
 }
+
+func TestGossipsubLimitIWANT(t *testing.T) {
+	ctx, cancel := context.WithCancel(context.Background())
+	defer cancel()
+	hosts := getDefaultHosts(t, 3)
+	denseConnect(t, hosts)
+
+	psubs := make([]*PubSub, 3)
+
+	defaultParams := DefaultGossipSubParams()
+	defaultParams.MaxIWantsPerMessageIDPerHeartbeat = 1
+	// Delay the heartbeat so we don't reset our count of allowed IWANTS for the
+	// first part of the test.
+	defaultParams.HeartbeatInitialDelay = 4 * time.Second
+	defaultParams.HeartbeatInterval = time.Second
+
+	psubs[0] = getGossipsub(ctx, hosts[0], WithGossipSubParams(defaultParams))
+
+	var iwantsRecvd atomic.Int32
+
+	copy(psubs[1:], getGossipsubs(ctx, hosts[1:], WithRawTracer(&mockRawTracer{
+		onRecvRPC: func(rpc *RPC) {
+			if len(rpc.GetControl().GetIwant()) > 0 {
+				iwantsRecvd.Add(1)
+			}
+		},
+	})))
+
+	topicString := "foobar"
+	for _, ps := range psubs {
+		topic, err := ps.Join(topicString)
+		if err != nil {
+			t.Fatal(err)
+		}
+
+		_, err = topic.Subscribe()
+		if err != nil {
+			t.Fatal(err)
+		}
+	}
+	time.Sleep(2 * time.Second)
+
+	publishIWant := func() {
+		psubs[1].eval <- func() {
+			psubs[1].rt.(*GossipSubRouter).sendRPC(hosts[0].ID(), &RPC{
+				RPC: pb.RPC{
+					Control: &pb.ControlMessage{
+						Ihave: []*pb.ControlIHave{
+							{
+								TopicID:    &topicString,
+								MessageIDs: []string{"1"},
+							},
+						},
+					},
+				},
+			}, false)
+		}
+
+		psubs[2].eval <- func() {
+			psubs[2].rt.(*GossipSubRouter).sendRPC(hosts[0].ID(), &RPC{
+				RPC: pb.RPC{
+					Control: &pb.ControlMessage{
+						Ihave: []*pb.ControlIHave{
+							{
+								TopicID:    &topicString,
+								MessageIDs: []string{"1"},
+							},
+						},
+					},
+				},
+			}, false)
+		}
+	}
+
+	publishIWant()
+	time.Sleep(time.Second)
+
+	if iwantsRecvd.Swap(0) != 1 {
+		t.Fatal("Expected exactly 1 IWANT due to limits")
+	}
+
+	// Wait for heartbeat to reset limit
+	time.Sleep(2 * time.Second)
+
+	publishIWant()
+	time.Sleep(time.Second)
+
+	if iwantsRecvd.Swap(0) != 1 {
+		t.Fatal("Expected exactly 1 IWANT due to limits")
+	}
+}