diff --git a/crates/polars-io/src/parquet/read/read_impl.rs b/crates/polars-io/src/parquet/read/read_impl.rs
index 93395a1324d9..ea867d34ddc5 100644
--- a/crates/polars-io/src/parquet/read/read_impl.rs
+++ b/crates/polars-io/src/parquet/read/read_impl.rs
@@ -350,16 +350,12 @@ fn rg_to_dfs_prefiltered(
         eprintln!("parquet live columns = {num_live_columns}, dead columns = {num_dead_columns}");
     }
 
-    // @NOTE: This is probably already sorted, but just to be sure.
-    let mut projection_sorted = projection.to_vec();
-    projection_sorted.sort();
-
     // We create two look-up tables that map indexes offsets into the live- and dead-set onto
     // column indexes of the schema.
     // Note: This may contain less than `num_live_columns` if there are hive columns involved.
     let mut live_idx_to_col_idx = Vec::with_capacity(num_live_columns);
-    let mut dead_idx_to_col_idx = Vec::with_capacity(num_dead_columns);
-    for &i in projection_sorted.iter() {
+    let mut dead_idx_to_col_idx: Vec<usize> = Vec::with_capacity(num_dead_columns);
+    for &i in projection.iter() {
         let name = schema.get_at_index(i).unwrap().0.as_str();
 
         if live_variables.contains(name) {
@@ -547,7 +543,7 @@ fn rg_to_dfs_prefiltered(
 
                 assert_eq!(
                     live_columns.len() + dead_columns.len(),
-                    projection_sorted.len() + hive_partition_columns.map_or(0, |x| x.len())
+                    projection.len() + hive_partition_columns.map_or(0, |x| x.len())
                 );
 
                 let mut merged = Vec::with_capacity(live_columns.len() + dead_columns.len());
diff --git a/py-polars/tests/unit/io/test_hive.py b/py-polars/tests/unit/io/test_hive.py
index 1bf018e969f2..f9f60afb209d 100644
--- a/py-polars/tests/unit/io/test_hive.py
+++ b/py-polars/tests/unit/io/test_hive.py
@@ -502,7 +502,13 @@ def test_hive_partition_force_async_17155(tmp_path: Path, monkeypatch: Any) -> N
     ("scan_func", "write_func"),
     [
         (partial(pl.scan_parquet, parallel="row_groups"), pl.DataFrame.write_parquet),
-        (partial(pl.scan_parquet, parallel="prefiltered"), pl.DataFrame.write_parquet),
+        (partial(pl.scan_parquet, parallel="columns"), pl.DataFrame.write_parquet),
+        (
+            lambda *a, **kw: pl.scan_parquet(*a, parallel="prefiltered", **kw).filter(
+                pl.col("b") == pl.col("b")
+            ),
+            pl.DataFrame.write_parquet,
+        ),
         (pl.scan_ipc, pl.DataFrame.write_ipc),
     ],
 )
diff --git a/py-polars/tests/unit/io/test_parquet.py b/py-polars/tests/unit/io/test_parquet.py
index cffc702be24a..adda7db9256c 100644
--- a/py-polars/tests/unit/io/test_parquet.py
+++ b/py-polars/tests/unit/io/test_parquet.py
@@ -2533,3 +2533,49 @@ def test_categorical_parametric_sliced(s: pl.Series, start: int, length: int) ->
             pl.scan_parquet(f).slice(start, length).collect(),
             df.slice(start, length),
         )
+
+
+@pytest.mark.write_disk
+def test_prefilter_with_projection_column_order_20175(tmp_path: Path) -> None:
+    path = tmp_path / "1"
+
+    pl.DataFrame({"a": 1, "b": 1, "c": 1, "d": 1, "e": 1}).write_parquet(path)
+
+    q = (
+        pl.scan_parquet(path, parallel="prefiltered")
+        .filter(pl.col("a") == 1)
+        .select("a", "d", "c")
+    )
+
+    assert_frame_equal(q.collect(), pl.DataFrame({"a": 1, "d": 1, "c": 1}))
+
+    f = io.BytesIO()
+
+    pl.read_csv(b"""\
+c0,c1,c2,c3,c4,c5,c6,c7,c8,c9,c10
+1,1,1,1,1,1,1,1,1,1,1
+1,1,1,1,1,1,1,1,1,1,1
+""").write_parquet(f)
+
+    f.seek(0)
+
+    q = (
+        pl.scan_parquet(
+            f,
+            rechunk=True,
+            parallel="prefiltered",
+        )
+        .filter(
+            pl.col("c0") == 1,
+        )
+        .select("c0", "c9", "c3")
+    )
+
+    assert_frame_equal(
+        q.collect(),
+        pl.read_csv(b"""\
+c0,c9,c3
+1,1,1
+1,1,1
+"""),
+    )