diff --git a/compiler/src/iree/compiler/InputConversion/StableHLO/Preprocessing/Canonicalization.cpp b/compiler/src/iree/compiler/InputConversion/StableHLO/Preprocessing/Canonicalization.cpp
index 07e95e8c7d9a..26cab680c34c 100644
--- a/compiler/src/iree/compiler/InputConversion/StableHLO/Preprocessing/Canonicalization.cpp
+++ b/compiler/src/iree/compiler/InputConversion/StableHLO/Preprocessing/Canonicalization.cpp
@@ -978,6 +978,31 @@ struct ReshapeOpCanon final : OpRewritePattern<mlir::stablehlo::ReshapeOp> {
   }
 };
 
+struct MergeConsecutiveReshapes final
+    : OpRewritePattern<mlir::stablehlo::ReshapeOp> {
+  using OpRewritePattern::OpRewritePattern;
+
+  LogicalResult matchAndRewrite(mlir::stablehlo::ReshapeOp op,
+                                PatternRewriter &rewriter) const override {
+    // Fold noop reshape.
+    auto operand = op.getOperand();
+    if (op.getType() == operand.getType()) {
+      rewriter.replaceOp(op, op.getOperand());
+      return success();
+    }
+
+    // Fold reshape(reshape(x)).
+    auto reshapeOp = operand.getDefiningOp<mlir::stablehlo::ReshapeOp>();
+    if (!reshapeOp) {
+      return rewriter.notifyMatchFailure(
+          op, "requires defining op of operand to be Reshape");
+    }
+
+    op.setOperand(reshapeOp->getOperand(0));
+    return success();
+  }
+};
+
 struct TransposeIsReshape final
     : OpRewritePattern<mlir::stablehlo::TransposeOp> {
   using OpRewritePattern::OpRewritePattern;
@@ -1156,7 +1181,7 @@ void populateCanonicalizationPatterns(MLIRContext *context,
       NoopReduceOpCanon, EmptyReduceOpCanon,
       // Shape manipulation(-ish) ops.
       ConcatenateOpCanon, ConvertOpCanon, DynamicReshapeOpCanon, GatherOpCanon,
-      ReshapeOpCanon, TransposeIsReshape,
+      ReshapeOpCanon, MergeConsecutiveReshapes, TransposeIsReshape,
       // Types.
       ZeroExtentTensorCanon>(context, benefit);
   patterns->add<ReorderElementwiseAndShapeOp>(context);
diff --git a/compiler/src/iree/compiler/InputConversion/StableHLO/Preprocessing/test/canonicalization.mlir b/compiler/src/iree/compiler/InputConversion/StableHLO/Preprocessing/test/canonicalization.mlir
index e8eb9825f321..af6df8e81fcb 100644
--- a/compiler/src/iree/compiler/InputConversion/StableHLO/Preprocessing/test/canonicalization.mlir
+++ b/compiler/src/iree/compiler/InputConversion/StableHLO/Preprocessing/test/canonicalization.mlir
@@ -401,6 +401,17 @@ func.func @reshape(%arg0: tensor<1xf32>)
 
 // -----
 
+// CHECK-LABEL: @merge_consecutive_reshapes
+// CHECK-SAME: %[[ARG0:[a-zA-Z0-9$._-]+]]
+func.func @merge_consecutive_reshapes(%arg0: tensor<4x4xi32>) -> tensor<16xi32> {
+  %0 = stablehlo.reshape %arg0 : (tensor<4x4xi32>) -> tensor<2x8xi32>
+  %1 = stablehlo.reshape %0 : (tensor<2x8xi32>) -> tensor<16xi32>
+  // CHECK: [[R0:%.+]] = stablehlo.reshape %[[ARG0]] : (tensor<4x4xi32>) -> tensor<16xi32>
+  return %1 : tensor<16xi32>
+}
+
+// -----
+
 // CHECK-LABEL: func.func @transpose
 // CHECK-SAME:   ([[ARG0:%.+]]: tensor<2xf32>, [[ARG1:%.+]]: tensor<3x2xf32>, [[ARG2:%.+]]: tensor<f32>)
 func.func @transpose(%arg0: tensor<2xf32>, %arg1: tensor<3x2xf32>, %arg2: tensor<f32>)