diff --git a/src/citra_qt/configuration/config.cpp b/src/citra_qt/configuration/config.cpp
index 74d779b0a9..58495ae0cd 100644
--- a/src/citra_qt/configuration/config.cpp
+++ b/src/citra_qt/configuration/config.cpp
@@ -452,6 +452,7 @@ void Config::ReadCoreValues() {
 
     if (global) {
         ReadBasicSetting(Settings::values.use_cpu_jit);
+        ReadBasicSetting(Settings::values.delay_start_for_lle_modules);
     }
 
     qt_config->endGroup();
@@ -979,6 +980,7 @@ void Config::SaveCoreValues() {
 
     if (global) {
         WriteBasicSetting(Settings::values.use_cpu_jit);
+        WriteBasicSetting(Settings::values.delay_start_for_lle_modules);
     }
 
     qt_config->endGroup();
diff --git a/src/citra_qt/configuration/configure_debug.cpp b/src/citra_qt/configuration/configure_debug.cpp
index 98d95c9e03..9143959e3b 100644
--- a/src/citra_qt/configuration/configure_debug.cpp
+++ b/src/citra_qt/configuration/configure_debug.cpp
@@ -94,6 +94,8 @@ void ConfigureDebug::SetConfiguration() {
     ui->toggle_console->setChecked(UISettings::values.show_console.GetValue());
     ui->log_filter_edit->setText(QString::fromStdString(Settings::values.log_filter.GetValue()));
     ui->toggle_cpu_jit->setChecked(Settings::values.use_cpu_jit.GetValue());
+    ui->delay_start_for_lle_modules->setChecked(
+        Settings::values.delay_start_for_lle_modules.GetValue());
     ui->toggle_renderer_debug->setChecked(Settings::values.renderer_debug.GetValue());
     ui->toggle_dump_command_buffers->setChecked(Settings::values.dump_command_buffers.GetValue());
 
@@ -125,6 +127,7 @@ void ConfigureDebug::ApplyConfiguration() {
     filter.ParseFilterString(Settings::values.log_filter.GetValue());
     Common::Log::SetGlobalFilter(filter);
     Settings::values.use_cpu_jit = ui->toggle_cpu_jit->isChecked();
+    Settings::values.delay_start_for_lle_modules = ui->delay_start_for_lle_modules->isChecked();
     Settings::values.renderer_debug = ui->toggle_renderer_debug->isChecked();
     Settings::values.dump_command_buffers = ui->toggle_dump_command_buffers->isChecked();
 
diff --git a/src/citra_qt/configuration/configure_debug.ui b/src/citra_qt/configuration/configure_debug.ui
index 20b84441cb..faf8b9ee0e 100644
--- a/src/citra_qt/configuration/configure_debug.ui
+++ b/src/citra_qt/configuration/configure_debug.ui
@@ -219,6 +219,25 @@
      </layout>
     </widget>
    </item>
+    <item>
+      <widget class="QGroupBox" name="groupBox_5">
+        <property name="title">
+          <string>Miscellaneus</string>
+        </property>
+        <layout class="QGridLayout" name="gridLayout_2">
+          <item row="1" column="0">
+            <widget class="QCheckBox" name="delay_start_for_lle_modules">
+              <property name="toolTip">
+                <string>&lt;html&gt;&lt;head/&gt;&lt;body&gt;&lt;p&gt;Introduces a delay to the first ever launched app thread if LLE modules are enabled, to allow them to initialize.&lt;/p&gt;&lt;/body&gt;&lt;/html&gt;</string>
+              </property>
+              <property name="text">
+                <string>Delay app start for LLE module initialization</string>
+              </property>
+            </widget>
+          </item>
+        </layout>
+      </widget>
+    </item>
    <item>
     <widget class="QLabel" name="label_cpu_clock_info">
      <property name="text">
diff --git a/src/common/settings.cpp b/src/common/settings.cpp
index 486340a3fe..e7e3d59f15 100644
--- a/src/common/settings.cpp
+++ b/src/common/settings.cpp
@@ -141,6 +141,7 @@ void LogSettings() {
     log_setting("System_RegionValue", values.region_value.GetValue());
     log_setting("System_PluginLoader", values.plugin_loader_enabled.GetValue());
     log_setting("System_PluginLoaderAllowed", values.allow_plugin_loader.GetValue());
+    log_setting("Debugging_DelayStartForLLEModules", values.delay_start_for_lle_modules.GetValue());
     log_setting("Debugging_UseGdbstub", values.use_gdbstub.GetValue());
     log_setting("Debugging_GdbstubPort", values.gdbstub_port.GetValue());
 }
diff --git a/src/common/settings.h b/src/common/settings.h
index 3b51e5dd45..8fc0f36c42 100644
--- a/src/common/settings.h
+++ b/src/common/settings.h
@@ -529,6 +529,7 @@ struct Values {
     // Debugging
     bool record_frame_times;
     std::unordered_map<std::string, bool> lle_modules;
+    Setting<bool> delay_start_for_lle_modules{true, "delay_start_for_lle_modules"};
     Setting<bool> use_gdbstub{false, "use_gdbstub"};
     Setting<u16> gdbstub_port{24689, "gdbstub_port"};
 
diff --git a/src/core/hle/kernel/kernel.cpp b/src/core/hle/kernel/kernel.cpp
index db0caf59c5..9df6e75ddb 100644
--- a/src/core/hle/kernel/kernel.cpp
+++ b/src/core/hle/kernel/kernel.cpp
@@ -183,6 +183,7 @@ void KernelSystem::serialize(Archive& ar, const unsigned int file_version) {
     ar& next_thread_id;
     ar& memory_mode;
     ar& n3ds_hw_caps;
+    ar& main_thread_extended_sleep;
     // Deliberately don't include debugger info to allow debugging through loads
 
     if (Archive::is_loading::value) {
diff --git a/src/core/hle/kernel/kernel.h b/src/core/hle/kernel/kernel.h
index e9f77be6c2..8c3f2b35c8 100644
--- a/src/core/hle/kernel/kernel.h
+++ b/src/core/hle/kernel/kernel.h
@@ -185,12 +185,14 @@ public:
      * @param processor_id The ID(s) of the processors on which the thread is desired to be run
      * @param stack_top The address of the thread's stack top
      * @param owner_process The parent process for the thread
+     * @param make_ready If the thread should be put in the ready queue
      * @return A shared pointer to the newly created thread
      */
     ResultVal<std::shared_ptr<Thread>> CreateThread(std::string name, VAddr entry_point,
                                                     u32 priority, u32 arg, s32 processor_id,
                                                     VAddr stack_top,
-                                                    std::shared_ptr<Process> owner_process);
+                                                    std::shared_ptr<Process> owner_process,
+                                                    bool make_ready = true);
 
     /**
      * Creates a semaphore.
@@ -338,6 +340,15 @@ public:
 
     Core::Timing& timing;
 
+    /// Sleep main thread of the first ever launched non-sysmodule process.
+    void SetAppMainThreadExtendedSleep(bool requires_sleep) {
+        main_thread_extended_sleep = requires_sleep;
+    }
+
+    bool GetAppMainThreadExtendedSleep() const {
+        return main_thread_extended_sleep;
+    }
+
 private:
     void MemoryInit(MemoryMode memory_mode, New3dsMemoryMode n3ds_mode, u64 override_init_time);
 
@@ -386,6 +397,14 @@ private:
      */
     std::recursive_mutex hle_lock;
 
+    /*
+     * Flags non system module main threads to wait a bit before running. On real hardware,
+     * system modules have plenty of time to load before the game is loaded, but on citra they
+     * start at the same time as the game. The artificial wait gives system modules some time
+     * to load and setup themselves before the game starts.
+     */
+    bool main_thread_extended_sleep = false;
+
     friend class boost::serialization::access;
     template <class Archive>
     void serialize(Archive& ar, const unsigned int file_version);
diff --git a/src/core/hle/kernel/thread.cpp b/src/core/hle/kernel/thread.cpp
index 83a2ba8f76..0bcca61772 100644
--- a/src/core/hle/kernel/thread.cpp
+++ b/src/core/hle/kernel/thread.cpp
@@ -10,8 +10,10 @@
 #include "common/common_types.h"
 #include "common/logging/log.h"
 #include "common/serialization/boost_flat_set.h"
+#include "common/settings.h"
 #include "core/arm/arm_interface.h"
 #include "core/arm/skyeye_common/armstate.h"
+#include "core/core.h"
 #include "core/hle/kernel/errors.h"
 #include "core/hle/kernel/kernel.h"
 #include "core/hle/kernel/mutex.h"
@@ -324,7 +326,7 @@ static void ResetThreadContext(Core::ARM_Interface::ThreadContext& context, u32
 
 ResultVal<std::shared_ptr<Thread>> KernelSystem::CreateThread(
     std::string name, VAddr entry_point, u32 priority, u32 arg, s32 processor_id, VAddr stack_top,
-    std::shared_ptr<Process> owner_process) {
+    std::shared_ptr<Process> owner_process, bool make_ready) {
     // Check if priority is in ranged. Lowest priority -> highest priority id.
     if (priority > ThreadPrioLowest) {
         LOG_ERROR(Kernel_SVC, "Invalid thread priority: {}", priority);
@@ -367,8 +369,11 @@ ResultVal<std::shared_ptr<Thread>> KernelSystem::CreateThread(
     // to initialize the context
     ResetThreadContext(thread->context, stack_top, entry_point, arg);
 
-    thread_managers[processor_id]->ready_queue.push_back(thread->current_priority, thread.get());
-    thread->status = ThreadStatus::Ready;
+    if (make_ready) {
+        thread_managers[processor_id]->ready_queue.push_back(thread->current_priority,
+                                                             thread.get());
+        thread->status = ThreadStatus::Ready;
+    }
 
     return thread;
 }
@@ -405,16 +410,36 @@ void Thread::BoostPriority(u32 priority) {
 
 std::shared_ptr<Thread> SetupMainThread(KernelSystem& kernel, u32 entry_point, u32 priority,
                                         std::shared_ptr<Process> owner_process) {
+
+    constexpr s64 sleep_app_thread_ns = 2'600'000'000LL;
+    constexpr u32 system_module_tid_high = 0x00040130;
+
+    const bool is_lle_service =
+        static_cast<u32>(owner_process->codeset->program_id >> 32) == system_module_tid_high;
+
+    s64 sleep_time_ns = 0;
+    if (!is_lle_service && kernel.GetAppMainThreadExtendedSleep()) {
+        if (Settings::values.delay_start_for_lle_modules) {
+            sleep_time_ns = sleep_app_thread_ns;
+        }
+        kernel.SetAppMainThreadExtendedSleep(false);
+    }
+
     // Initialize new "main" thread
     auto thread_res =
         kernel.CreateThread("main", entry_point, priority, 0, owner_process->ideal_processor,
-                            Memory::HEAP_VADDR_END, owner_process);
+                            Memory::HEAP_VADDR_END, owner_process, sleep_time_ns == 0);
 
     std::shared_ptr<Thread> thread = std::move(thread_res).Unwrap();
 
     thread->context.fpscr =
         FPSCR_DEFAULT_NAN | FPSCR_FLUSH_TO_ZERO | FPSCR_ROUND_TOZERO | FPSCR_IXC; // 0x03C00010
 
+    if (sleep_time_ns != 0) {
+        thread->status = ThreadStatus::WaitSleep;
+        thread->WakeAfterDelay(sleep_time_ns);
+    }
+
     // Note: The newly created thread will be run when the scheduler fires.
     return thread;
 }
diff --git a/src/core/hle/service/service.cpp b/src/core/hle/service/service.cpp
index 118ea32dd1..133bfce209 100644
--- a/src/core/hle/service/service.cpp
+++ b/src/core/hle/service/service.cpp
@@ -214,10 +214,20 @@ static bool AttemptLLE(const ServiceModuleInfo& service_module) {
 /// Initialize ServiceManager
 void Init(Core::System& core) {
     SM::ServiceManager::InstallInterfaces(core);
+    core.Kernel().SetAppMainThreadExtendedSleep(false);
+    bool lle_module_present = false;
 
     for (const auto& service_module : service_module_map) {
-        if (!AttemptLLE(service_module) && service_module.init_function != nullptr)
+        const bool has_lle = AttemptLLE(service_module);
+        if (!has_lle && service_module.init_function != nullptr) {
             service_module.init_function(core);
+        }
+        lle_module_present |= has_lle;
+    }
+    if (lle_module_present) {
+        // If there is at least one LLE module, tell the kernel to
+        // add a extended sleep to the app main thread (if option enabled).
+        core.Kernel().SetAppMainThreadExtendedSleep(true);
     }
     LOG_DEBUG(Service, "initialized OK");
 }