diff --git a/src/mito2/src/read/scan_region.rs b/src/mito2/src/read/scan_region.rs
index 40f18c393b..d28562c554 100644
--- a/src/mito2/src/read/scan_region.rs
+++ b/src/mito2/src/read/scan_region.rs
@@ -14,6 +14,7 @@
 
 //! Scans a region according to the scan request.
 
+use std::collections::HashSet;
 use std::fmt;
 use std::sync::Arc;
 use std::time::{Duration, Instant};
@@ -24,6 +25,7 @@ use common_telemetry::{debug, error, tracing, warn};
 use common_time::range::TimestampRange;
 use common_time::Timestamp;
 use datafusion::physical_plan::DisplayFormatType;
+use datafusion_expr::utils::expr_to_columns;
 use smallvec::SmallVec;
 use store_api::region_engine::{PartitionRange, RegionScannerRef};
 use store_api::storage::{ScanRequest, TimeSeriesRowSelector};
@@ -295,6 +297,9 @@ impl ScanRegion {
             self.version.options.append_mode,
         );
 
+        // Remove field filters for LastNonNull mode after logging the request.
+        self.maybe_remove_field_filters();
+
         let inverted_index_applier = self.build_invereted_index_applier();
         let fulltext_index_applier = self.build_fulltext_index_applier();
         let predicate = Predicate::new(self.request.filters.clone());
@@ -321,7 +326,7 @@ impl ScanRegion {
         Ok(input)
     }
 
-    /// Build time range predicate from filters.
+    /// Build time range predicate from filters, also remove time filters from request.
     fn build_time_range_predicate(&mut self) -> TimestampRange {
         let time_index = self.version.metadata.time_index_column();
         let unit = time_index
@@ -337,6 +342,38 @@ impl ScanRegion {
         )
     }
 
+    /// Remove field filters if the merge mode is [MergeMode::LastNonNull].
+    fn maybe_remove_field_filters(&mut self) {
+        if self.version.options.merge_mode() != MergeMode::LastNonNull {
+            return;
+        }
+
+        // TODO(yingwen): We can ignore field filters only when there are multiple sources in the same time window.
+        let field_columns = self
+            .version
+            .metadata
+            .field_columns()
+            .map(|col| &col.column_schema.name)
+            .collect::<HashSet<_>>();
+        // Columns in the expr.
+        let mut columns = HashSet::new();
+
+        self.request.filters.retain(|expr| {
+            columns.clear();
+            // `expr_to_columns` won't return error.
+            if expr_to_columns(expr, &mut columns).is_err() {
+                return false;
+            }
+            for column in &columns {
+                if field_columns.contains(&column.name) {
+                    // This expr uses the field column.
+                    return false;
+                }
+            }
+            true
+        });
+    }
+
     /// Use the latest schema to build the inveretd index applier.
     fn build_invereted_index_applier(&self) -> Option<InvertedIndexApplierRef> {
         if self.ignore_inverted_index {
diff --git a/tests/cases/standalone/common/select/prune_field.result b/tests/cases/standalone/common/select/prune_field.result
new file mode 100644
index 0000000000..d309d384be
--- /dev/null
+++ b/tests/cases/standalone/common/select/prune_field.result
@@ -0,0 +1,69 @@
+CREATE TABLE IF NOT EXISTS prune_field (
+  ts TIMESTAMP TIME INDEX,
+  tag UInt16,
+  a UInt8,
+  b UInt8,
+PRIMARY KEY (tag)) ENGINE = mito WITH('merge_mode'='last_non_null');
+
+Affected Rows: 0
+
+insert into prune_field(ts, tag, a, b) values(0, 1, 1, null);
+
+Affected Rows: 1
+
+admin flush_table('prune_field');
+
++----------------------------------+
+| ADMIN flush_table('prune_field') |
++----------------------------------+
+| 0                                |
++----------------------------------+
+
+insert into prune_field(ts, tag, a, b) values(0, 1, null, 1);
+
+Affected Rows: 1
+
+admin flush_table('prune_field');
+
++----------------------------------+
+| ADMIN flush_table('prune_field') |
++----------------------------------+
+| 0                                |
++----------------------------------+
+
+select * from prune_field where a = 1;
+
++---------------------+-----+---+---+
+| ts                  | tag | a | b |
++---------------------+-----+---+---+
+| 1970-01-01T00:00:00 | 1   | 1 | 1 |
++---------------------+-----+---+---+
+
+select * from prune_field where b = 1;
+
++---------------------+-----+---+---+
+| ts                  | tag | a | b |
++---------------------+-----+---+---+
+| 1970-01-01T00:00:00 | 1   | 1 | 1 |
++---------------------+-----+---+---+
+
+select * from prune_field;
+
++---------------------+-----+---+---+
+| ts                  | tag | a | b |
++---------------------+-----+---+---+
+| 1970-01-01T00:00:00 | 1   | 1 | 1 |
++---------------------+-----+---+---+
+
+select * from prune_field where a = 1 and b = 1;
+
++---------------------+-----+---+---+
+| ts                  | tag | a | b |
++---------------------+-----+---+---+
+| 1970-01-01T00:00:00 | 1   | 1 | 1 |
++---------------------+-----+---+---+
+
+drop table prune_field;
+
+Affected Rows: 0
+
diff --git a/tests/cases/standalone/common/select/prune_field.sql b/tests/cases/standalone/common/select/prune_field.sql
new file mode 100644
index 0000000000..6488175508
--- /dev/null
+++ b/tests/cases/standalone/common/select/prune_field.sql
@@ -0,0 +1,24 @@
+CREATE TABLE IF NOT EXISTS prune_field (
+  ts TIMESTAMP TIME INDEX,
+  tag UInt16,
+  a UInt8,
+  b UInt8,
+PRIMARY KEY (tag)) ENGINE = mito WITH('merge_mode'='last_non_null');
+
+insert into prune_field(ts, tag, a, b) values(0, 1, 1, null);
+
+admin flush_table('prune_field');
+
+insert into prune_field(ts, tag, a, b) values(0, 1, null, 1);
+
+admin flush_table('prune_field');
+
+select * from prune_field where a = 1;
+
+select * from prune_field where b = 1;
+
+select * from prune_field;
+
+select * from prune_field where a = 1 and b = 1;
+
+drop table prune_field;