diff --git a/F3DEX2.cpp b/F3DEX2.cpp
index d902c51b..a865dcf3 100644
--- a/F3DEX2.cpp
+++ b/F3DEX2.cpp
@@ -201,6 +201,7 @@ void F3DEX2_DMAIO( u32 w0, u32 w1 )
 
 void F3DEX2_Special_1( u32 w0, u32 w1 )
 {
+	gSPDlistCount(_SHIFTR( w0, 0, 8 ), w1);
 }
 
 void F3DEX2_Special_2( u32 w0, u32 w1 )
diff --git a/RSP.cpp b/RSP.cpp
index 5b347f93..6f94962b 100644
--- a/RSP.cpp
+++ b/RSP.cpp
@@ -121,6 +121,18 @@ LoadLoop:
 #endif // WIN32_ASM
 }
 
+void RSP_CheckDLCounter()
+{
+	if (RSP.count != -1) {
+		--RSP.count;
+		if (RSP.count == 0) {
+			RSP.count = -1;
+			--RSP.PCi;
+			DebugMsg( DEBUG_LOW | DEBUG_HANDLED, "End of DL\n" );
+		}
+	}
+}
+
 void RSP_ProcessDList()
 {
 	VI_UpdateSize();
@@ -128,7 +140,7 @@ void RSP_ProcessDList()
 
 	RSP.PC[0] = *(u32*)&DMEM[0x0FF0];
 	RSP.PCi = 0;
-	RSP.count = 0;
+	RSP.count = -1;
 
 	RSP.halt = FALSE;
 	RSP.busy = TRUE;
@@ -213,6 +225,7 @@ void RSP_ProcessDList()
 		RSP.nextCmd = _SHIFTR( *(u32*)&RDRAM[RSP.PC[RSP.PCi]], 24, 8 );
 
 		GBI.cmd[RSP.cmd]( w0, w1 );
+		RSP_CheckDLCounter();
 	}
 
 	if (config.frameBufferEmulation.copyToRDRAM)
diff --git a/RSP.h b/RSP.h
index c4a84e3a..795a5222 100644
--- a/RSP.h
+++ b/RSP.h
@@ -5,7 +5,8 @@
 
 typedef struct
 {
-	u32 PC[18], PCi, busy, halt, close, DList, uc_start, uc_dstart, cmd, nextCmd, count;
+	u32 PC[18], PCi, busy, halt, close, DList, uc_start, uc_dstart, cmd, nextCmd;
+	s32 count;
 } RSPInfo;
 
 extern RSPInfo RSP;
@@ -15,5 +16,6 @@ extern RSPInfo RSP;
 void RSP_Init();
 void RSP_ProcessDList();
 void RSP_LoadMatrix( f32 mtx[4][4], u32 address );
+void RSP_CheckDLCounter();
 
 #endif
diff --git a/gSP.cpp b/gSP.cpp
index b7e9d043..3c93c661 100644
--- a/gSP.cpp
+++ b/gSP.cpp
@@ -934,6 +934,29 @@ void gSPBranchLessZ( u32 branchdl, u32 vtx, f32 zval )
 #endif
 }
 
+void gSPDlistCount(u32 count, u32 v)
+{
+	u32 address = RSP_SegmentToPhysical( v );
+	if (address == 0 || (address + 8) > RDRAMSize) {
+		DebugMsg( DEBUG_HIGH | DEBUG_ERROR, "// Attempting to branch to display list at invalid address\n" );
+		DebugMsg( DEBUG_HIGH | DEBUG_HANDLED, "gSPDlistCnt(%d, 0x%08X );\n", count, v );
+		return;
+	}
+
+	if (RSP.PCi >= 9) {
+		DebugMsg( DEBUG_HIGH | DEBUG_ERROR, "// ** DL stack overflow **\n" );
+		DebugMsg( DEBUG_HIGH | DEBUG_HANDLED, "gSPDlistCnt(%d, 0x%08X );\n", count, v );
+		return;
+	}
+
+	DebugMsg( DEBUG_HIGH | DEBUG_HANDLED, "gSPDlistCnt(%d, 0x%08X );\n", count, v );
+
+	++RSP.PCi;  // go to the next PC in the stack
+	RSP.PC[RSP.PCi] = address;  // jump to the address
+	RSP.nextCmd = _SHIFTR( *(u32*)&RDRAM[address], 24, 8 );
+	RSP.count = count + 1;
+}
+
 void gSPSetDMAOffsets( u32 mtxoffset, u32 vtxoffset )
 {
 	gSP.DMAOffsets.mtx = mtxoffset;
diff --git a/gSP.h b/gSP.h
index b5e6105f..77a89bbb 100644
--- a/gSP.h
+++ b/gSP.h
@@ -166,6 +166,7 @@ void gSPDisplayList( u32 dl );
 void gSPDMADisplayList( u32 dl, u32 n );
 void gSPBranchList( u32 dl );
 void gSPBranchLessZ( u32 branchdl, u32 vtx, f32 zval );
+void gSPDlistCount(u32 count, u32 v);
 void gSPSprite2DBase( u32 base );
 void gSPDMATriangles( u32 tris, u32 n );
 void gSP1Quadrangle( s32 v0, s32 v1, s32 v2, s32 v3 );