diff --git a/src/controller/myapp.c b/src/controller/myapp.c
index fffb8d7..838bad5 100644
--- a/src/controller/myapp.c
+++ b/src/controller/myapp.c
@@ -211,6 +211,37 @@ int myapp_run(myconf_t *myconf)
     uint64_t my_lsn = (uint64_t)atomic_load_explicit(&c.ha.applied_lsn,
                                                        memory_order_acquire);
 
+    /* Startup stagger: delay election start by priority * election_priority_step_ms.
+     * Node 0 starts first; lower-priority nodes wait, see the winner's vote request
+     * arrive, and grant the vote instead of starting their own election.
+     * Uses the same step as the existing blocking-election probe path (Req 1.2). */
+    if (c.ha.my_priority > 0) {
+        uint32_t step_ms = c.ha.config.election_priority_step_ms
+                           ? c.ha.config.election_priority_step_ms
+                           : HA_DEFAULT_ELECTION_PRIORITY_STEP_MS;
+        long delay_ns = (long)c.ha.my_priority * (long)step_ms * 1000000L;
+        struct timespec jitter = {
+            .tv_sec  = delay_ns / 1000000000L,
+            .tv_nsec = delay_ns % 1000000000L,
+        };
+        nanosleep(&jitter, NULL);
+    }
+
+    /* If we already voted for someone during the startup delay (incoming vote
+     * request arrived from the higher-priority node that started first), skip
+     * starting our own election — we'll become STANDBY via the election timeout
+     * watchdog on the winner's side, or promote ourselves if no heartbeat arrives. */
+    {
+        pthread_mutex_lock(&c.ha.lock);
+        int already_voted = (c.ha.voted_for_name[0] != '\0');
+        pthread_mutex_unlock(&c.ha.lock);
+        if (already_voted) {
+            INFO("myapp_run: already voted during startup delay — "
+                 "skipping own election, waiting for primary");
+            goto skip_election;
+        }
+    }
+
     /* Non-blocking: arms election timeout watchdog, connects outgoing vote
      * endpoints, returns immediately.  Election resolves via iomgr callbacks;
      * ha_elect_evaluate_and_signal calls cluster_init_phase2 from vtask context
@@ -219,6 +250,8 @@ int myapp_run(myconf_t *myconf)
                             (n_other > 0) ? peer_ptrs : NULL,
                             my_tag, my_lsn, myconf->name);
 
+skip_election:;
+
     /* ── Step 9: cluster_run ─────────────────────────────────────────────── */
 
     /* Req 9.4: myapp_run exits into cluster_run within 10ms of startup.