diff --git a/fml/thread.cc b/fml/thread.cc
index cff08b4f8f87e..8fc43592d06d3 100644
--- a/fml/thread.cc
+++ b/fml/thread.cc
@@ -108,7 +108,11 @@ void SetThreadName(const std::string& name) {
 #if defined(FML_OS_MACOSX)
   pthread_setname_np(name.c_str());
 #elif defined(FML_OS_LINUX) || defined(FML_OS_ANDROID)
-  pthread_setname_np(pthread_self(), name.c_str());
+  // Linux thread names are limited to 16 characters including the terminating
+  // null.
+  constexpr std::string::size_type kLinuxMaxThreadNameLen = 15;
+  pthread_setname_np(pthread_self(),
+                     name.substr(0, kLinuxMaxThreadNameLen).c_str());
 #elif defined(FML_OS_WIN)
   THREADNAME_INFO info;
   info.dwType = 0x1000;
diff --git a/fml/thread_unittests.cc b/fml/thread_unittests.cc
index 0f668aed3ed5d..8e596edee4618 100644
--- a/fml/thread_unittests.cc
+++ b/fml/thread_unittests.cc
@@ -149,4 +149,19 @@ TEST(Thread, ThreadPriorityCreatedWithConfig) {
   thread.Join();
   ASSERT_TRUE(done);
 }
-#endif
+#endif  // FLUTTER_PTHREAD_SUPPORTED
+
+#if defined(FML_OS_LINUX)
+TEST(Thread, LinuxLongThreadNameTruncated) {
+  const std::string name = "VeryLongThreadNameTest";
+  fml::Thread thread(name);
+
+  thread.GetTaskRunner()->PostTask([&name]() {
+    constexpr size_t kThreadNameLen = 16;
+    char thread_name[kThreadNameLen];
+    pthread_getname_np(pthread_self(), thread_name, kThreadNameLen);
+    ASSERT_EQ(thread_name, name.substr(0, kThreadNameLen - 1));
+  });
+  thread.Join();
+}
+#endif  // FML_OS_LINUX
diff --git a/impeller/renderer/backend/vulkan/context_vk.cc b/impeller/renderer/backend/vulkan/context_vk.cc
index 6c1804dd74022..df1dea8e795f0 100644
--- a/impeller/renderer/backend/vulkan/context_vk.cc
+++ b/impeller/renderer/backend/vulkan/context_vk.cc
@@ -130,7 +130,7 @@ void ContextVK::Setup(Settings settings) {
     return;
   }
 
-  queue_submit_thread_ = std::make_unique<fml::Thread>("QueueSubmitThread");
+  queue_submit_thread_ = std::make_unique<fml::Thread>("IplrVkQueueSub");
   queue_submit_thread_->GetTaskRunner()->PostTask([]() {
     // submitKHR is extremely cheap and mostly blocks on an internal fence.
     fml::RequestAffinity(fml::CpuAffinity::kEfficiency);
diff --git a/impeller/renderer/backend/vulkan/fence_waiter_vk.cc b/impeller/renderer/backend/vulkan/fence_waiter_vk.cc
index 11ee7f0b7db72..a43426f452fe6 100644
--- a/impeller/renderer/backend/vulkan/fence_waiter_vk.cc
+++ b/impeller/renderer/backend/vulkan/fence_waiter_vk.cc
@@ -91,7 +91,7 @@ static std::vector<vk::Fence> GetFencesForWaitSet(const WaitSet& set) {
 
 void FenceWaiterVK::Main() {
   fml::Thread::SetCurrentThreadName(
-      fml::Thread::ThreadConfig{"io.flutter.impeller.fence_waiter"});
+      fml::Thread::ThreadConfig{"IplrVkFenceWait"});
   // Since this thread mostly waits on fences, it doesn't need to be fast.
   fml::RequestAffinity(fml::CpuAffinity::kEfficiency);
 
diff --git a/impeller/renderer/backend/vulkan/resource_manager_vk.cc b/impeller/renderer/backend/vulkan/resource_manager_vk.cc
index b719737337f8e..8b55801047ec9 100644
--- a/impeller/renderer/backend/vulkan/resource_manager_vk.cc
+++ b/impeller/renderer/backend/vulkan/resource_manager_vk.cc
@@ -38,8 +38,7 @@ void ResourceManagerVK::Start() {
   //
   // ... so no FML_DCHECK here.
 
-  fml::Thread::SetCurrentThreadName(
-      fml::Thread::ThreadConfig{"io.flutter.impeller.resource_manager"});
+  fml::Thread::SetCurrentThreadName(fml::Thread::ThreadConfig{"IplrVkResMgr"});
   // While this code calls destructors it doesn't need to be particularly fast
   // with them, as long as it doesn't interrupt raster thread.
   fml::RequestAffinity(fml::CpuAffinity::kEfficiency);