diff --git a/project/common.mk b/project/common.mk
index fe3a444..5854810 100644
--- a/project/common.mk
+++ b/project/common.mk
@@ -33,6 +33,7 @@ COMMON_SRCS = \
 	src/process/ntapi_tt_create_native_process_v2.c \
 	src/process/ntapi_tt_create_remote_process_params.c \
 	src/process/ntapi_tt_create_remote_runtime_data.c \
+	src/process/ntapi_tt_fork.c \
 	src/process/ntapi_tt_fork_v1.c \
 	src/process/ntapi_tt_fork_v2.c \
 	src/process/ntapi_tt_get_runtime_data.c \
diff --git a/src/internal/ntapi.c b/src/internal/ntapi.c
index 8ebdf22..fe928ac 100644
--- a/src/internal/ntapi.c
+++ b/src/internal/ntapi.c
@@ -211,6 +211,7 @@ static int32_t __fastcall __ntapi_init_once(ntapi_vtbl ** pvtbl)
 	__ntapi->tt_create_thread 				= __ntapi_tt_create_thread;
 
 	/* nt_process.h */
+	__ntapi->tt_fork					= __ntapi_tt_fork;
 	__ntapi->tt_create_remote_process_params 		= __ntapi_tt_create_remote_process_params;
 	__ntapi->tt_spawn_native_process			= __ntapi_tt_spawn_native_process;
 	__ntapi->tt_get_runtime_data 				= __ntapi_tt_get_runtime_data;
@@ -361,7 +362,6 @@ static int32_t __fastcall __ntapi_init_once(ntapi_vtbl ** pvtbl)
 
 	/* OS version dependent functions */
 	if (__ntapi->zw_create_user_process) {
-		__ntapi->tt_fork 				= __ntapi_tt_fork_v2;
 		__ntapi->tt_create_native_process		= __ntapi_tt_create_native_process_v2;
 		__ntapi->ipc_create_pipe 			= __ntapi_ipc_create_pipe_v2;
 		__ntapi->sc_socket				= __ntapi_sc_socket_v2;
@@ -371,7 +371,6 @@ static int32_t __fastcall __ntapi_init_once(ntapi_vtbl ** pvtbl)
 		__ntapi->sc_getsockname				= __ntapi_sc_getsockname_v2;
 		__ntapi->sc_getpeername				= __ntapi_sc_getpeername_v2;
 	} else {
-		__ntapi->tt_fork = __ntapi_tt_fork_v1;
 		__ntapi->tt_create_native_process		= __ntapi_tt_create_native_process_v1;
 		__ntapi->ipc_create_pipe 			= __ntapi_ipc_create_pipe_v1;
 		__ntapi->sc_socket				= __ntapi_sc_socket_v1;
diff --git a/src/internal/ntapi_fnapi.h b/src/internal/ntapi_fnapi.h
index ee291ff..650839e 100644
--- a/src/internal/ntapi_fnapi.h
+++ b/src/internal/ntapi_fnapi.h
@@ -94,6 +94,7 @@ ntapi_tt_create_local_thread			__ntapi_tt_create_local_thread;
 ntapi_tt_create_remote_thread			__ntapi_tt_create_remote_thread;
 
 /* nt_process.h */
+ntapi_tt_fork					__ntapi_tt_fork;
 ntapi_tt_fork					__ntapi_tt_fork_v1;
 ntapi_tt_fork					__ntapi_tt_fork_v2;
 ntapi_tt_create_remote_process_params		__ntapi_tt_create_remote_process_params;
diff --git a/src/process/ntapi_tt_fork.c b/src/process/ntapi_tt_fork.c
new file mode 100644
index 0000000..9f498eb
--- /dev/null
+++ b/src/process/ntapi_tt_fork.c
@@ -0,0 +1,230 @@
+/********************************************************/
+/*  ntapi: Native API core library                      */
+/*  Copyright (C) 2013--2016  Z. Gilboa                 */
+/*  Released under GPLv2 and GPLv3; see COPYING.NTAPI.  */
+/********************************************************/
+
+#include <psxtypes/psxtypes.h>
+#include <ntapi/nt_atomic.h>
+#include <ntapi/nt_status.h>
+#include <ntapi/nt_object.h>
+#include <ntapi/nt_memory.h>
+#include <ntapi/nt_thread.h>
+#include <ntapi/nt_process.h>
+#include <ntapi/ntapi.h>
+#include "ntapi_impl.h"
+
+static intptr_t	__fork_retry_stats  = 0;
+static intptr_t	__fork_resume_stats = 0;
+
+static int32_t __stdcall __fork_thread(void * ctx)
+{
+	intptr_t *	pstate;
+	intptr_t	state;
+	void *		hready;
+
+	pstate = (intptr_t *)ctx;
+	state  = *pstate;
+	hready = (void *)state;
+
+	at_store(
+		pstate,
+		0);
+
+	return __ntapi->zw_terminate_thread(
+		NT_CURRENT_THREAD_HANDLE,
+		__ntapi->zw_set_event(
+			hready,0));
+}
+
+static intptr_t __fastcall __ntapi_tt_fork_child(
+	void *		hresumed,
+	void *		hready)
+{
+	int32_t			status;
+	nt_thread_params	tparams;
+	nt_timeout		timeout;
+	nt_timeout		zerowait;
+	intptr_t		state;
+
+	at_store(
+		&state,
+		(intptr_t)hready);
+
+	__ntapi->tt_aligned_block_memset(
+		&tparams,0,sizeof(tparams));
+
+	tparams.start			= __fork_thread;
+	tparams.arg			= &state;
+	tparams.stack_size_commit	= 0x10000;
+	tparams.stack_size_reserve	= 0x20000;
+
+	status = __ntapi->tt_create_local_thread(
+		&tparams);
+
+	__ntapi->zw_set_event(
+		hresumed,0);
+
+	if (status)
+		__ntapi->zw_terminate_process(
+			NT_CURRENT_PROCESS_HANDLE,
+			status);
+
+	if (!state) {
+		__ntapi->zw_close(hresumed);
+		__ntapi->zw_close(hready);
+		__ntapi->zw_close(tparams.hthread);
+		return 0;
+	}
+
+	timeout.quad  = (-1) * 10 * 1000 * 250;
+
+	status = __ntapi->zw_wait_for_single_object(
+		hready,
+		NT_SYNC_NON_ALERTABLE,
+		&timeout);
+
+	if (status == NT_STATUS_SUCCESS) {
+		__ntapi->zw_close(hresumed);
+		__ntapi->zw_close(hready);
+		__ntapi->zw_close(tparams.hthread);
+		return 0;
+	}
+
+	__ntapi->zw_terminate_thread(
+		tparams.hthread,
+		NT_STATUS_MORE_PROCESSING_REQUIRED);
+
+	zerowait.quad = 0;
+
+	status = __ntapi->zw_wait_for_single_object(
+		hready,
+		NT_SYNC_NON_ALERTABLE,
+		&zerowait);
+
+	if (status == NT_STATUS_SUCCESS) {
+		__ntapi->zw_close(hresumed);
+		__ntapi->zw_close(hready);
+		__ntapi->zw_close(tparams.hthread);
+		return 0;
+	}
+
+	return __ntapi->zw_terminate_process(
+		NT_CURRENT_PROCESS_HANDLE,
+		status);
+}
+
+static intptr_t __fastcall __ntapi_tt_fork_parent(
+	void **		hprocess,
+	void **		hthread,
+	void *		hresumed,
+	void *		hready)
+{
+	int32_t		status;
+	nt_timeout	timeout;
+	nt_timeout	zerowait;
+	uint32_t	prev;
+
+	__ntapi->zw_wait_for_single_object(
+		hresumed,
+		NT_SYNC_NON_ALERTABLE,
+		0);
+
+	timeout.quad = (-1) * 10 * 1000 * 500;
+
+	status = __ntapi->zw_wait_for_single_object(
+		hready,
+		NT_SYNC_NON_ALERTABLE,
+		&timeout);
+
+	if (status == NT_STATUS_SUCCESS) {
+		__ntapi->zw_close(hresumed);
+		__ntapi->zw_close(hready);
+		return NT_STATUS_SUCCESS;
+	}
+
+	__ntapi->zw_suspend_thread(
+		*hthread,&prev);
+
+	zerowait.quad = 0;
+
+	status = __ntapi->zw_wait_for_single_object(
+		hready,
+		NT_SYNC_NON_ALERTABLE,
+		&zerowait);
+
+	if (status == NT_STATUS_SUCCESS) {
+		at_locked_inc(
+			&__fork_resume_stats);
+
+		__ntapi->zw_resume_thread(
+			*hthread,0);
+
+		__ntapi->zw_close(hresumed);
+		__ntapi->zw_close(hready);
+		return NT_STATUS_SUCCESS;
+	}
+
+	at_locked_inc(
+		&__fork_retry_stats);
+
+	__ntapi->zw_terminate_process(
+		*hprocess,
+		status);
+
+	__ntapi->zw_close(*hprocess);
+	__ntapi->zw_close(*hthread);
+
+	return status;
+}
+
+intptr_t __fastcall __ntapi_tt_fork(
+	__out	void **		hprocess,
+	__out	void **		hthread)
+{
+	int32_t			status;
+	intptr_t		pid;
+	void *			hresumed;
+	void *			hready;
+	int			i;
+
+	if ((status = __ntapi->tt_create_inheritable_event(
+			&hresumed,
+			NT_NOTIFICATION_EVENT,
+			NT_EVENT_NOT_SIGNALED)))
+		return -1;
+
+	if ((status = __ntapi->tt_create_inheritable_event(
+			&hready,
+			NT_NOTIFICATION_EVENT,
+			NT_EVENT_NOT_SIGNALED)))
+		return -1;
+
+	for (i=0; i<32; i++) {
+		if (__ntapi->zw_create_user_process)
+			pid = __ntapi_tt_fork_v2(hprocess,hthread);
+		else
+			pid = __ntapi_tt_fork_v2(hprocess,hthread);
+
+		if (pid == 0) {
+			return __ntapi_tt_fork_child(
+				hresumed,hready);
+
+		} else if (pid > 0) {
+			if (!(__ntapi_tt_fork_parent(
+					hprocess,hthread,
+					hresumed,hready)))
+				return pid;
+
+		} else {
+			__ntapi->zw_close(hresumed);
+			__ntapi->zw_close(hready);
+			return -1;
+		}
+	}
+
+	__ntapi->zw_close(hresumed);
+	__ntapi->zw_close(hready);
+
+	return -1;
+}