diff --git a/GPU/Common/VertexDecoderCommon.h b/GPU/Common/VertexDecoderCommon.h
index fbe60c175698..2066defced53 100644
--- a/GPU/Common/VertexDecoderCommon.h
+++ b/GPU/Common/VertexDecoderCommon.h
@@ -26,6 +26,8 @@
 #include "Core/Reporting.h"
 #include "GPU/ge_constants.h"
 #include "GPU/Common/ShaderCommon.h"
+#include "GPU/GPUCommon.h"
+
 #if PPSSPP_ARCH(ARM)
 #include "Common/ArmEmitter.h"
 #elif PPSSPP_ARCH(ARM64)
@@ -78,30 +80,6 @@ struct DecVtxFormat {
 	void ComputeID();
 };
 
-struct TransformedVertex
-{
-	union {
-		struct {
-			float x, y, z, fog;     // in case of morph, preblend during decode
-		};
-		float pos[4];
-	};
-	union {
-		struct {
-			float u; float v; float w;   // scaled by uscale, vscale, if there
-		};
-		float uv[3];
-	};
-	union {
-		u8 color0[4];   // prelit
-		u32 color0_32;
-	};
-	union {
-		u8 color1[4];   // prelit
-		u32 color1_32;
-	};
-};
-
 void GetIndexBounds(const void *inds, int count, u32 vertType, u16 *indexLowerBound, u16 *indexUpperBound);
 
 inline int RoundUp4(int x) {
diff --git a/GPU/GPUCommon.cpp b/GPU/GPUCommon.cpp
index 4a836550f183..23d9577725e8 100644
--- a/GPU/GPUCommon.cpp
+++ b/GPU/GPUCommon.cpp
@@ -303,16 +303,16 @@ const CommonCommandTableEntry commonCommandTable[] = {
 	{ GE_CMD_BONEMATRIXDATA,    FLAG_EXECUTE, 0, &GPUCommon::Execute_BoneMtxData },
 
 	// Vertex Screen/Texture/Color
-	{ GE_CMD_VSCX, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
-	{ GE_CMD_VSCY, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
-	{ GE_CMD_VSCZ, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
-	{ GE_CMD_VTCS, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
-	{ GE_CMD_VTCT, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
-	{ GE_CMD_VTCQ, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
-	{ GE_CMD_VCV, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
-	{ GE_CMD_VAP, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
-	{ GE_CMD_VFC, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
-	{ GE_CMD_VSCV, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
+	{ GE_CMD_VSCX },
+	{ GE_CMD_VSCY },
+	{ GE_CMD_VSCZ },
+	{ GE_CMD_VTCS },
+	{ GE_CMD_VTCT },
+	{ GE_CMD_VTCQ },
+	{ GE_CMD_VCV },
+	{ GE_CMD_VAP, FLAG_EXECUTE, 0, &GPUCommon::Execute_ImmVertexAlphaPrim },
+	{ GE_CMD_VFC },
+	{ GE_CMD_VSCV },
 
 	// "Missing" commands (gaps in the sequence)
 	{ GE_CMD_UNKNOWN_03, FLAG_EXECUTE, 0, &GPUCommon::Execute_Unknown },
@@ -1808,6 +1808,32 @@ void GPUCommon::Execute_MorphWeight(u32 op, u32 diff) {
 	gstate_c.morphWeights[(op >> 24) - GE_CMD_MORPHWEIGHT0] = getFloat24(op);
 }
 
+void GPUCommon::Execute_ImmVertexAlphaPrim(u32 op, u32 diff) {
+	uint32_t data = op & 0xFFFFFF;
+	TransformedVertex &v = immBuffer_[immCount_++];
+
+	// Formula deduced from ThrillVille's clear.
+	int offsetX = gstate.getOffsetX16();
+	int offsetY = gstate.getOffsetY16();
+	v.x = ((gstate.imm_vscx & 0xFFFFFF) - offsetX) / 16.0f;
+	v.y = ((gstate.imm_vscy & 0xFFFFFF) - offsetY) / 16.0f;
+	v.z = gstate.imm_vscz & 0xFFFF;
+	v.u = 0.0f;  // we have no information about the scale here
+	v.v = 0.0f;  // we have no information about the scale here
+	v.w = 0.0f;  // we have no information about the scale here
+	v.color0_32 = (gstate.imm_cv & 0xFFFFFF) | (gstate.imm_ap << 24);
+	v.fog = 0.0f; // we have no information about the scale here
+	v.color1_32 = gstate.imm_scv & 0xFFFFFF;
+	int prim = (op >> 8) & 0xF;
+	if (prim != 7) {
+		immPrim_ = prim;
+	} else {
+		// Time to submit! This can go through the software transform path but skipping the actual transform...
+		// drawEngineCommon_->SubmitImm(immBuffer_, immCount_);
+		immCount_ = 0;
+	}
+}
+
 void GPUCommon::ExecuteOp(u32 op, u32 diff) {
 	const u32 cmd = op >> 24;
 
diff --git a/GPU/GPUCommon.h b/GPU/GPUCommon.h
index c4d22dc904da..41db2a9f3f08 100644
--- a/GPU/GPUCommon.h
+++ b/GPU/GPUCommon.h
@@ -39,6 +39,29 @@ enum {
 	FLAG_DIRTYONCHANGE = 64,  // NOTE: Either this or FLAG_EXECUTE*, not both!
 };
 
+struct TransformedVertex {
+	union {
+		struct {
+			float x, y, z, fog;     // in case of morph, preblend during decode
+		};
+		float pos[4];
+	};
+	union {
+		struct {
+			float u; float v; float w;   // scaled by uscale, vscale, if there
+		};
+		float uv[3];
+	};
+	union {
+		u8 color0[4];   // prelit
+		u32 color0_32;
+	};
+	union {
+		u8 color1[4];   // prelit
+		u32 color1_32;
+	};
+};
+
 class GPUCommon : public GPUInterface, public GPUDebugInterface {
 public:
 	GPUCommon(GraphicsContext *gfxCtx, Draw::DrawContext *draw);
@@ -124,6 +147,8 @@ class GPUCommon : public GPUInterface, public GPUDebugInterface {
 
 	void Execute_MorphWeight(u32 op, u32 diff);
 
+	void Execute_ImmVertexAlphaPrim(u32 op, u32 diff);
+
 	void Execute_Unknown(u32 op, u32 diff);
 
 	int EstimatePerVertexCost();
@@ -291,6 +316,15 @@ class GPUCommon : public GPUInterface, public GPUDebugInterface {
 	DrawType lastDraw_;
 	GEPrimitiveType lastPrim_;
 
+	// No idea how big this buffer needs to be.
+	enum {
+		MAX_IMMBUFFER_SIZE = 32,
+	};
+
+	TransformedVertex immBuffer_[MAX_IMMBUFFER_SIZE];
+	int immCount_ = 0;
+	int immPrim_ = 0;
+
 private:
 	// Debug stats.
 	double timeSteppingStarted_;
diff --git a/GPU/GPUState.h b/GPU/GPUState.h
index a78aa10307e2..cd41c10a76d1 100644
--- a/GPU/GPUState.h
+++ b/GPU/GPUState.h
@@ -185,8 +185,18 @@ struct GPUgstate {
 				transfersrcpos,
 				transferdstpos,
 				pad99,
-				transfersize;  // 0xEE
-
+				transfersize,  // 0xEE
+				pad100,         // 0xEF
+				imm_vscx,        // 0xF0
+				imm_vscy,
+				imm_vscz,
+				imm_vtcs,
+				imm_vtct,
+				imm_vtcq,
+				imm_cv,
+				imm_ap,
+				imm_fc,
+				imm_scv;
 			u32 pad05[0xFF- 0xEE];
 		};
 	};