diff --git a/tests/ThreadPool_Tests.cpp b/tests/ThreadPool_Tests.cpp
new file mode 100644
index 00000000..7e84d894
--- /dev/null
+++ b/tests/ThreadPool_Tests.cpp
@@ -0,0 +1,165 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Digital Voice Modem - Test Suite
+ * GPLv2 Open Source. Use is subject to license terms.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ *  Copyright (C) 2026 Bryan Biedenkapp, N2PLL
+ *
+ */
+#include "Defines.h"
+#include "common/ThreadPool.h"
+
+#include <catch2/catch_test_macros.hpp>
+
+#include <atomic>
+#include <chrono>
+#include <condition_variable>
+#include <mutex>
+
+// ---------------------------------------------------------------------------
+//  Global Functions
+// ---------------------------------------------------------------------------
+
+/**
+ * @brief Waits until the atomic value reaches the target or the timeout expires.
+ * @param value The atomic value to monitor.
+ * @param target The target value to wait for.
+ * @param timeoutMs The maximum time to wait in milliseconds.
+ * @return true if the target value was reached, false if the timeout expired.
+ */
+bool waitForCount(std::atomic<uint32_t>& value, uint32_t target, uint32_t timeoutMs = 1500U)
+{
+    const auto deadline = std::chrono::steady_clock::now() + std::chrono::milliseconds(timeoutMs);
+    while (std::chrono::steady_clock::now() < deadline) {
+        if (value.load() >= target)
+            return true;
+
+        Thread::sleep(1U);
+    }
+
+    return value.load() >= target;
+}
+
+TEST_CASE("ThreadPool enforces minimum worker count", "[common][threadpool]")
+{
+    ThreadPool pool(1U, "tp-min");
+    REQUIRE(pool.getMaxWorkerCnt() == 4U);
+}
+
+TEST_CASE("ThreadPool runs enqueued tasks", "[common][threadpool]")
+{
+    ThreadPool pool(4U, "tp-run");
+    std::atomic<uint32_t> executed { 0U };
+
+    pool.start();
+
+    const uint32_t taskCount = 12U;
+    bool allAccepted = true;
+    for (uint32_t i = 0U; i < taskCount; i++) {
+        ThreadPoolTask* task = new_pooltask([&executed]() {
+            executed.fetch_add(1U);
+        });
+
+        if (!pool.enqueue(task)) {
+            delete task;
+            allAccepted = false;
+            break;
+        }
+    }
+    REQUIRE(allAccepted);
+    REQUIRE(waitForCount(executed, taskCount));
+
+    pool.stop();
+    pool.wait();
+
+    REQUIRE(executed.load() == taskCount);
+}
+
+TEST_CASE("ThreadPool rejects enqueue when stopped", "[common][threadpool]")
+{
+    ThreadPool pool(4U, "tp-stop");
+
+    ThreadPoolTask* beforeStart = new_pooltask([]() {});
+    const bool beforeStartAccepted = pool.enqueue(beforeStart);
+    REQUIRE_FALSE(beforeStartAccepted);
+    if (!beforeStartAccepted)
+        delete beforeStart;
+
+    pool.start();
+    pool.stop();
+
+    ThreadPoolTask* afterStop = new_pooltask([]() {});
+    const bool afterStopAccepted = pool.enqueue(afterStop);
+    REQUIRE_FALSE(afterStopAccepted);
+    if (!afterStopAccepted)
+        delete afterStop;
+
+    pool.wait();
+}
+
+TEST_CASE("ThreadPool max queued task limit is enforced", "[common][threadpool]")
+{
+    ThreadPool pool(4U, "tp-queue");
+    pool.setMaxQueuedTasks(0U);
+
+    std::atomic<uint32_t> executed { 0U };
+    std::mutex gateMutex;
+    std::condition_variable gateCond;
+    bool releaseWorkers = false;
+
+    pool.start();
+
+    // Block all workers so one extra task remains queued.
+    bool workerTasksAccepted = true;
+    for (uint32_t i = 0U; i < pool.getMaxWorkerCnt(); i++) {
+        ThreadPoolTask* task = new_pooltask([&]() {
+            std::unique_lock<std::mutex> lock(gateMutex);
+            gateCond.wait(lock, [&]() { return releaseWorkers; });
+            executed.fetch_add(1U);
+        });
+
+        if (!pool.enqueue(task)) {
+            delete task;
+            workerTasksAccepted = false;
+            break;
+        }
+    }
+
+    if (workerTasksAccepted)
+        Thread::sleep(10U);
+
+    // Now enforce a one-item queue while all workers are occupied.
+    pool.setMaxQueuedTasks(1U);
+
+    ThreadPoolTask* acceptedQueuedTask = new_pooltask([&executed]() {
+        executed.fetch_add(1U);
+    });
+    bool queuedAccepted = false;
+    if (workerTasksAccepted)
+        queuedAccepted = pool.enqueue(acceptedQueuedTask);
+    if (!queuedAccepted)
+        delete acceptedQueuedTask;
+
+    ThreadPoolTask* overflowTask = new_pooltask([]() {});
+    bool overflowAccepted = false;
+    if (workerTasksAccepted && queuedAccepted)
+        overflowAccepted = pool.enqueue(overflowTask);
+    if (!overflowAccepted)
+        delete overflowTask;
+
+    // scope is intentional
+    {
+        std::lock_guard<std::mutex> lock(gateMutex);
+        releaseWorkers = true;
+    }
+    gateCond.notify_all();
+
+    REQUIRE(workerTasksAccepted);
+    REQUIRE(queuedAccepted);
+    REQUIRE_FALSE(overflowAccepted);
+    REQUIRE(waitForCount(executed, pool.getMaxWorkerCnt()));
+
+    pool.stop();
+    pool.wait();
+}
\ No newline at end of file
diff --git a/tests/Thread_Tests.cpp b/tests/Thread_Tests.cpp
new file mode 100644
index 00000000..f207d677
--- /dev/null
+++ b/tests/Thread_Tests.cpp
@@ -0,0 +1,206 @@
+// SPDX-License-Identifier: GPL-2.0-only
+/*
+ * Digital Voice Modem - Test Suite
+ * GPLv2 Open Source. Use is subject to license terms.
+ * DO NOT ALTER OR REMOVE COPYRIGHT NOTICES OR THIS FILE HEADER.
+ *
+ *  Copyright (C) 2026 Bryan Biedenkapp, N2PLL
+ *
+ */
+#include "Defines.h"
+#include "common/Thread.h"
+
+#include <catch2/catch_test_macros.hpp>
+
+#include <atomic>
+#include <chrono>
+#include <condition_variable>
+#include <mutex>
+
+// ---------------------------------------------------------------------------
+//  Class Declaration
+// ---------------------------------------------------------------------------
+
+/**
+ * @brief Test thread implementation for unit tests.
+ */
+class TestThread final : public Thread {
+public:
+    /**
+     * @brief Constructs a TestThread instance.
+     * @param entryCount The atomic counter to track thread entry.
+     * @param mutex The mutex to protect shared state.
+     * @param cond The condition variable to signal thread completion.
+     * @param ran The flag indicating whether the thread has run.
+     */
+    TestThread(std::atomic<uint32_t>& entryCount, std::mutex& mutex,
+        std::condition_variable& cond, bool& ran) :
+        m_entryCount(entryCount),
+        m_mutex(mutex),
+        m_cond(cond),
+        m_ran(ran)
+    {
+        /* stub */
+    }
+
+    /**
+     * @brief Thread entry point.
+     */
+    void entry() override
+    {
+        m_entryCount.fetch_add(1U);
+
+        // scope is intentional
+        {
+            std::lock_guard<std::mutex> lock(m_mutex);
+            m_ran = true;
+        }
+
+        m_cond.notify_one();
+    }
+
+private:
+    std::atomic<uint32_t>& m_entryCount;
+    std::mutex& m_mutex;
+    std::condition_variable& m_cond;
+    bool& m_ran;
+};
+
+// ---------------------------------------------------------------------------
+//  Structure Declaration
+// ---------------------------------------------------------------------------
+
+/**
+ * @brief Context structure for running a thread routine.
+ */
+struct RunAsThreadContext {
+    std::atomic<uint32_t> callCount { 0U };
+    std::mutex mutex;
+    std::condition_variable cond;
+    bool ran = false;
+};
+
+// ---------------------------------------------------------------------------
+//  Global Functions
+// ---------------------------------------------------------------------------
+
+/**
+ * @brief Thread start routine that executes the thread's entry function.
+ * @param arg Pointer to the thread_t structure.
+ * @return void* Always returns nullptr.
+ */
+void* runAsThreadRoutine(void* arg)
+{
+    thread_t* thread = static_cast<thread_t*>(arg);
+    if (thread == nullptr || thread->obj == nullptr)
+        return nullptr;
+
+    RunAsThreadContext* ctx = static_cast<RunAsThreadContext*>(thread->obj);
+    ctx->callCount.fetch_add(1U);
+
+    // scope is intentional
+    {
+        std::lock_guard<std::mutex> lock(ctx->mutex);
+        ctx->ran = true;
+    }
+
+    ctx->cond.notify_one();
+    return nullptr;
+}
+
+/**
+ * @brief Waits for a flag to become true with a timeout.
+ * @param mutex The mutex protecting the flag.
+ * @param cond The condition variable to wait on.
+ * @param flag The flag to wait for.
+ * @param timeoutMs The timeout in milliseconds.
+ * @return bool True if the flag became true within the timeout, false otherwise.
+ */
+bool waitForFlag(std::mutex& mutex, std::condition_variable& cond, bool& flag, uint32_t timeoutMs = 1000U)
+{
+    std::unique_lock<std::mutex> lock(mutex);
+    return cond.wait_for(lock, std::chrono::milliseconds(timeoutMs), [&flag]() { return flag; });
+}
+
+TEST_CASE("Thread runs entry once and wait joins", "[common][thread]")
+{
+    std::atomic<uint32_t> entryCount { 0U };
+    std::mutex mutex;
+    std::condition_variable cond;
+    bool ran = false;
+
+    TestThread thread(entryCount, mutex, cond, ran);
+
+    REQUIRE_FALSE(thread.started());
+    REQUIRE(thread.run());
+    REQUIRE(thread.started());
+
+    REQUIRE(waitForFlag(mutex, cond, ran));
+
+    // A second run() call should not spawn another worker.
+    REQUIRE(thread.run());
+    thread.wait();
+
+    REQUIRE(entryCount.load() == 1U);
+}
+
+TEST_CASE("Thread wait and setName are safe before run", "[common][thread]")
+{
+    std::atomic<uint32_t> entryCount { 0U };
+    std::mutex mutex;
+    std::condition_variable cond;
+    bool ran = false;
+
+    TestThread thread(entryCount, mutex, cond, ran);
+
+    thread.setName("not-running");
+    thread.wait();
+    thread.detach();
+
+    REQUIRE(entryCount.load() == 0U);
+    REQUIRE_FALSE(thread.started());
+}
+
+TEST_CASE("Thread detach allows completion without wait", "[common][thread]")
+{
+    std::atomic<uint32_t> entryCount { 0U };
+    std::mutex mutex;
+    std::condition_variable cond;
+    bool ran = false;
+
+    TestThread thread(entryCount, mutex, cond, ran);
+
+    REQUIRE(thread.run());
+    thread.detach();
+
+    REQUIRE(waitForFlag(mutex, cond, ran));
+    REQUIRE(entryCount.load() == 1U);
+}
+
+TEST_CASE("Thread runAsThread executes start routine", "[common][thread]")
+{
+    RunAsThreadContext ctx;
+    thread_t threadData;
+
+    REQUIRE(Thread::runAsThread(&ctx, runAsThreadRoutine, &threadData));
+    REQUIRE(waitForFlag(ctx.mutex, ctx.cond, ctx.ran));
+
+#if defined(_WIN32)
+    ::WaitForSingleObject(threadData.thread, INFINITE);
+    ::CloseHandle(threadData.thread);
+#else
+    ::pthread_join(threadData.thread, nullptr);
+#endif // defined(_WIN32)
+
+    REQUIRE(ctx.callCount.load() == 1U);
+}
+
+TEST_CASE("Thread sleep delays execution", "[common][thread]")
+{
+    const auto begin = std::chrono::steady_clock::now();
+    Thread::sleep(10U);
+    const auto end = std::chrono::steady_clock::now();
+
+    const auto elapsedMs = std::chrono::duration_cast<std::chrono::milliseconds>(end - begin).count();
+    REQUIRE(elapsedMs >= 5);
+}
\ No newline at end of file