diff --git a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.h b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.h
--- a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.h
+++ b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.h
@@ -25,6 +25,11 @@
 /// transformed to `shape.reduce`, which can be lowered to SCF and Standard.
 std::unique_ptr<Pass> createShapeToShapeLowering();
 
+/// Creates an instance of the ExpandReduce pass that rewrites Shape ReduceOp as
+/// an SCF ForOp. The operations within the region of ReduceOp are cloned to the
+/// body of ForOp.
+std::unique_ptr<Pass> createExpandReducePass();
+
 } // end namespace mlir
 
 #endif // MLIR_DIALECT_SHAPE_TRANSFORMS_PASSES_H_
diff --git a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
--- a/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
+++ b/mlir/include/mlir/Dialect/Shape/Transforms/Passes.td
@@ -11,9 +11,14 @@
 
 include "mlir/Pass/PassBase.td"
 
-def ShapeToShapeLowering : FunctionPass<"shape-to-shape-lowering"> {
+def ShapeToShapeLowering : FunctionPass<"shape-to-shape"> {
   let summary = "Legalize Shape dialect to be convertible to Standard";
   let constructor = "mlir::createShapeToShapeLowering()";
 }
 
+def ExpandReduce : FunctionPass<"expand-reduce"> {
+  let summary = "Rewrites `shape.reduce` as an `scf.for` loop";
+  let constructor = "mlir::createExpandReducePass()";
+}
+
 #endif // MLIR_DIALECT_SHAPE_TRANSFORMS_PASSES
diff --git a/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt b/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
--- a/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
+++ b/mlir/lib/Dialect/Shape/Transforms/CMakeLists.txt
@@ -1,4 +1,5 @@
 add_mlir_dialect_library(MLIRShapeOpsTransforms
+  ExpandReduce.cpp
   ShapeToShapeLowering.cpp
 
   ADDITIONAL_HEADER_DIRS
diff --git a/mlir/lib/Dialect/Shape/Transforms/ExpandReduce.cpp b/mlir/lib/Dialect/Shape/Transforms/ExpandReduce.cpp
new file mode 100644
--- /dev/null
+++ b/mlir/lib/Dialect/Shape/Transforms/ExpandReduce.cpp
@@ -0,0 +1,94 @@
+//===- ExpandReduce.cpp - Expand to SCF ForOp -----------------------------===//
+//
+// Part of the LLVM Project, under the Apache License v2.0 with LLVM Exceptions.
+// See https://llvm.org/LICENSE.txt for license information.
+// SPDX-License-Identifier: Apache-2.0 WITH LLVM-exception
+//
+//===----------------------------------------------------------------------===//
+
+#include "PassDetail.h"
+#include "mlir/Dialect/SCF/SCF.h"
+#include "mlir/Dialect/Shape/IR/Shape.h"
+#include "mlir/Dialect/Shape/Transforms/Passes.h"
+#include "mlir/Dialect/StandardOps/IR/Ops.h"
+#include "mlir/IR/BlockAndValueMapping.h"
+#include "mlir/IR/Builders.h"
+#include "mlir/IR/PatternMatch.h"
+#include "mlir/Pass/Pass.h"
+#include "mlir/Transforms/DialectConversion.h"
+
+using namespace mlir;
+using namespace mlir::shape;
+
+namespace {
+/// Converts `shape.reduce` to `scf.for`.
+struct ReduceOpConverter : public OpRewritePattern<ReduceOp> {
+public:
+  using OpRewritePattern::OpRewritePattern;
+
+  LogicalResult matchAndRewrite(ReduceOp op,
+                                PatternRewriter &rewriter) const final;
+};
+} // namespace
+
+LogicalResult
+ReduceOpConverter::matchAndRewrite(ReduceOp reduceOp,
+                                   PatternRewriter &rewriter) const {
+  auto loc = reduceOp.getLoc();
+
+  Value zero = rewriter.create<ConstantIndexOp>(loc, 0);
+  Value one = rewriter.create<ConstantIndexOp>(loc, 1);
+  Value extent_tensor = rewriter.create<ToExtentTensorOp>(
+      loc,
+      RankedTensorType::get({ShapedType::kDynamicSize},
+                            rewriter.getIndexType()),
+      reduceOp.shape());
+  Value rank = rewriter.create<RankOp>(loc, extent_tensor);
+
+  auto loop = rewriter.create<scf::ForOp>(
+      loc, zero, rank, one, reduceOp.initVals(),
+      [&](OpBuilder &b, Location nestedLoc, Value iv, ValueRange args) {
+        Value index_extent = b.create<ExtractElementOp>(loc, extent_tensor, iv);
+        Value size_extent = rewriter.create<IndexToSizeOp>(loc, index_extent);
+
+        SmallVector<Value, 2> mapped_values{iv, size_extent};
+        mapped_values.append(args.begin(), args.end());
+
+        BlockAndValueMapping mapping;
+        Block *loopBody = reduceOp.getBody();
+        mapping.map(loopBody->getArguments(), mapped_values);
+        for (auto &nested : loopBody->without_terminator()) {
+          Operation *clone = b.clone(nested, mapping);
+          mapping.map(nested.getResults(), clone->getResults());
+        }
+
+        SmallVector<Value, 2> mapped_results;
+        for (auto result : loopBody->getTerminator()->getOperands())
+          mapped_results.push_back(mapping.lookup(result));
+        b.create<scf::YieldOp>(loc, mapped_results);
+      });
+
+  rewriter.replaceOp(reduceOp, loop.getResults());
+  return success();
+}
+
+namespace {
+struct ExpandReduce : public ExpandReduceBase<ExpandReduce> {
+  void runOnFunction() override;
+};
+} // namespace
+
+void ExpandReduce::runOnFunction() {
+  OwningRewritePatternList patterns;
+  patterns.insert<ReduceOpConverter>(&getContext());
+
+  ConversionTarget target(getContext());
+  target.addLegalDialect<ShapeDialect, scf::SCFDialect, StandardOpsDialect>();
+  target.addIllegalOp<ReduceOp>();
+  if (failed(mlir::applyPartialConversion(getFunction(), target, patterns)))
+    signalPassFailure();
+}
+
+std::unique_ptr<Pass> mlir::createExpandReducePass() {
+  return std::make_unique<ExpandReduce>();
+}
diff --git a/mlir/test/Dialect/Shape/expand-reduce.mlir b/mlir/test/Dialect/Shape/expand-reduce.mlir
new file mode 100644
--- /dev/null
+++ b/mlir/test/Dialect/Shape/expand-reduce.mlir
@@ -0,0 +1,28 @@
+// RUN: mlir-opt -expand-reduce -split-input-file %s | FileCheck %s
+
+// CHECK-LABEL: shape_reduce
+// CHECK-SAME:   [[SHAPE:%.*]]: !shape.shape) -> !shape.size {
+func @shape_reduce(%shape : !shape.shape) -> !shape.size {
+  %init = shape.const_size 1
+  %num_elements = shape.reduce(%shape, %init) -> !shape.size {
+    ^bb0(%index: index, %dim: !shape.size, %acc: !shape.size):
+      %new_acc = shape.mul %acc, %dim
+      shape.yield %new_acc : !shape.size
+  }
+  return %num_elements : !shape.size
+}
+// CHECK-NEXT: [[SHAPE_C1:%.*]] = shape.const_size 1
+// CHECK-NEXT: [[C0:%.*]] = constant 0 : index
+// CHECK-NEXT: [[C1:%.*]] = constant 1 : index
+
+// CHECK-NEXT: [[EXTENTS:%.*]] = "shape.to_extent_tensor"([[SHAPE]])
+// CHECK-NEXT: [[RANK:%.*]] = rank [[EXTENTS]] : tensor<?xindex>
+
+// CHECK-NEXT: [[RESULT:%.*]] = scf.for [[I:%.*]] = [[C0]] to [[RANK]]
+// CHECK-SAME:       step [[C1]] iter_args([[ACC:%.*]] = [[SHAPE_C1]])
+// CHECK-NEXT:   [[EXTENT_INDEX:%.*]] = extract_element [[EXTENTS]]{{\[}}[[I]]]
+// CHECK-NEXT:   [[EXTENT:%.*]] = shape.index_to_size [[EXTENT_INDEX]]
+// CHECK-NEXT:   [[NEW_ACC:%.*]] = shape.mul [[ACC]], [[EXTENT]]
+// CHECK-NEXT:   scf.yield [[NEW_ACC]] : !shape.size
+// CHECK-NEXT: }
+// CHECK-NEXT: return [[RESULT]] : !shape.size
diff --git a/mlir/test/Dialect/Shape/shape-to-shape.mlir b/mlir/test/Dialect/Shape/shape-to-shape.mlir
--- a/mlir/test/Dialect/Shape/shape-to-shape.mlir
+++ b/mlir/test/Dialect/Shape/shape-to-shape.mlir
@@ -1,4 +1,4 @@
-// RUN: mlir-opt -shape-to-shape-lowering -split-input-file %s | FileCheck %s
+// RUN: mlir-opt -shape-to-shape -split-input-file %s | FileCheck %s
 
 // CHECK-LABEL: func @num_elements_to_reduce(
 // CHECK-SAME:    [[ARG:%.*]]: !shape.shape) -> [[SIZE_TY:!.*]] {