diff --git a/src/pdf_extract.py b/src/pdf_extract.py
index 31cac7b..1989043 100644
--- a/src/pdf_extract.py
+++ b/src/pdf_extract.py
@@ -778,9 +778,21 @@ def scan_pdf_for_transactions(
     metadata = extract_statement_metadata(pages)
 
     out_rows: list[dict[str, Any]] = []
-    prev: dict[str, Any] | None = None
+    # Maximum y-gap (in PDF points) between a transaction and a
+    # following no-date-no-amount line for that line to count as a
+    # continuation of the description. Typical line baselines sit
+    # ~10–14 pts apart; 25 pts allows for one blank line but
+    # rejects section headers that are several rows away.
+    _MULTILINE_MERGE_MAX_GAP = 25.0
 
     for page in pages:
+        # ``prev`` and ``prev_y_bottom`` reset per page so a section
+        # header at the top of page 2 can't attach to the last
+        # transaction on page 1 — PDF y-coordinates restart at the
+        # top of each page so the y-distance check is meaningless
+        # across page boundaries.
+        prev: dict[str, Any] | None = None
+        prev_y_bottom: float | None = None
         rows = cluster_rows(page.words, y_tolerance=y_tolerance)
         for row_words in rows:
             line = " ".join(w.text for w in row_words).strip()
@@ -791,18 +803,29 @@ def scan_pdf_for_transactions(
             amount_tokens = _find_amount_tokens(row_words)
 
             if not dates or not amount_tokens:
-                # Continuation candidate — a line on a transaction
-                # page that has neither a date nor an amount of its
-                # own. Attach to the previous row's description.
+                # Continuation candidate — a line with no date AND
+                # no amount of its own. Only attach to the previous
+                # transaction if (a) we have one, (b) it's on this
+                # same page, and (c) the y-gap to it is small enough
+                # that a human would read this as a wrapped line
+                # rather than a separate paragraph or section header.
                 if (
                     merge_multiline_descriptions
                     and prev is not None
                     and not dates
                     and not amount_tokens
+                    and row_words
                 ):
-                    prev["description"] = (
-                        (prev["description"] + " " + line).strip()
-                    )
+                    current_top = min(w.top for w in row_words)
+                    if (
+                        prev_y_bottom is not None
+                        and (current_top - prev_y_bottom)
+                        <= _MULTILINE_MERGE_MAX_GAP
+                    ):
+                        prev["description"] = (
+                            (prev["description"] + " " + line).strip()
+                        )
+                        prev_y_bottom = max(w.bottom for w in row_words)
                 continue
 
             # First date wins for the "date" column; ALL dates are
@@ -816,6 +839,14 @@ def scan_pdf_for_transactions(
                 row_words, date_ranges, amount_idxs,
             )
 
+            # Every real transaction must have a description. Rows
+            # like "01/13/2025  $1,000.00" (Daily Ledger Balances
+            # section, page totals, period summaries) carry a date
+            # and an amount but no text in between — they're
+            # statement furniture, not transactions. Drop them.
+            if not desc.strip():
+                continue
+
             iso = parse_date(first_date_text, date_formats)
             if iso is None:
                 # Short date — try to bind to the statement period
@@ -867,6 +898,9 @@ def scan_pdf_for_transactions(
 
             out_rows.append(record)
             prev = record
+            prev_y_bottom = (
+                max(w.bottom for w in row_words) if row_words else None
+            )
 
     return out_rows, warnings
 
diff --git a/tests/test_pdf_extract_smoke.py b/tests/test_pdf_extract_smoke.py
index 79a0d13..2baca94 100644
--- a/tests/test_pdf_extract_smoke.py
+++ b/tests/test_pdf_extract_smoke.py
@@ -341,6 +341,153 @@ class TestZeroAmountRowsAreDropped:
         assert rows[0]["amount_1"] == -40.00
 
 
+class TestRequiresDescription:
+    """Every kept row must have non-empty description. Filters out
+    "Daily Ledger Balances" entries (date + amount with no
+    description) and similar statement-furniture rows."""
+
+    def test_empty_description_row_dropped(self):
+        from src import pdf_extract as mod
+        from src.pdf_extract import Page, WordBox, scan_pdf_for_transactions
+
+        original = mod.extract_pages_auto
+
+        def fake(_b, *, allow_ocr=True):
+            words = [
+                # Real transaction
+                WordBox(x0=0, top=0, x1=80, bottom=10, text="01/13/2026"),
+                WordBox(x0=100, top=0, x1=160, bottom=10, text="Coffee"),
+                WordBox(x0=200, top=0, x1=240, bottom=10, text="$4.50"),
+                # Daily-balance row: date + amount, NO description
+                WordBox(x0=0, top=200, x1=80, bottom=210, text="01/14/2026"),
+                WordBox(x0=200, top=200, x1=280, bottom=210, text="$1,000.00"),
+            ]
+            return [Page(
+                page_no=1, width=300, height=300, text="", words=words,
+            )], []
+
+        mod.extract_pages_auto = fake
+        try:
+            rows, _ = scan_pdf_for_transactions(b"")
+        finally:
+            mod.extract_pages_auto = original
+
+        assert len(rows) == 1
+        assert "Coffee" in rows[0]["description"]
+
+
+class TestPrevTransactionResetsPerPage:
+    """A no-date no-amount line at the top of page 2 must NOT
+    attach to the last transaction of page 1. Different pages have
+    independent y-coordinate origins so the gap check would be
+    meaningless across pages."""
+
+    def test_no_cross_page_merge(self):
+        from src import pdf_extract as mod
+        from src.pdf_extract import Page, WordBox, scan_pdf_for_transactions
+
+        original = mod.extract_pages_auto
+
+        def fake(_b, *, allow_ocr=True):
+            p1 = Page(
+                page_no=1, width=300, height=300, text="",
+                words=[
+                    WordBox(x0=0, top=0, x1=80, bottom=10, text="01/13/2026"),
+                    WordBox(x0=100, top=0, x1=160, bottom=10, text="Coffee"),
+                    WordBox(x0=200, top=0, x1=240, bottom=10, text="$4.50"),
+                ],
+            )
+            p2 = Page(
+                page_no=2, width=300, height=300, text="",
+                # First content on page 2 is a section header with
+                # no date and no amount — should NOT attach to the
+                # Coffee row from page 1.
+                words=[
+                    WordBox(x0=0, top=0, x1=200, bottom=10, text="Daily"),
+                    WordBox(x0=50, top=0, x1=160, bottom=10, text="Ledger"),
+                    WordBox(x0=120, top=0, x1=240, bottom=10, text="Balances"),
+                ],
+            )
+            return [p1, p2], []
+
+        mod.extract_pages_auto = fake
+        try:
+            rows, _ = scan_pdf_for_transactions(b"")
+        finally:
+            mod.extract_pages_auto = original
+
+        assert len(rows) == 1
+        assert rows[0]["description"] == "Coffee"
+        assert "Ledger" not in rows[0]["description"]
+
+
+class TestMultilineMergeYGap:
+    """Wrapped-description continuations are close to the previous
+    transaction (~12 pts gap). Section headers further down the
+    page must NOT be silently merged in."""
+
+    def test_close_continuation_merges(self):
+        from src import pdf_extract as mod
+        from src.pdf_extract import Page, WordBox, scan_pdf_for_transactions
+
+        original = mod.extract_pages_auto
+
+        def fake(_b, *, allow_ocr=True):
+            words = [
+                # Transaction at y=0..10
+                WordBox(x0=0, top=0, x1=80, bottom=10, text="01/13/2026"),
+                WordBox(x0=100, top=0, x1=160, bottom=10, text="Coffee"),
+                WordBox(x0=200, top=0, x1=240, bottom=10, text="$4.50"),
+                # Continuation at y=20..30 (gap = 10 pts) — should merge
+                WordBox(x0=100, top=20, x1=200, bottom=30, text="Vendor"),
+                WordBox(x0=210, top=20, x1=260, bottom=30, text="memo"),
+            ]
+            return [Page(
+                page_no=1, width=300, height=50, text="", words=words,
+            )], []
+
+        mod.extract_pages_auto = fake
+        try:
+            rows, _ = scan_pdf_for_transactions(b"")
+        finally:
+            mod.extract_pages_auto = original
+
+        assert len(rows) == 1
+        assert "Vendor memo" in rows[0]["description"]
+
+    def test_far_section_header_does_not_merge(self):
+        """Same setup but the second line is far below — would be
+        a different paragraph in the source PDF."""
+        from src import pdf_extract as mod
+        from src.pdf_extract import Page, WordBox, scan_pdf_for_transactions
+
+        original = mod.extract_pages_auto
+
+        def fake(_b, *, allow_ocr=True):
+            words = [
+                WordBox(x0=0, top=0, x1=80, bottom=10, text="01/13/2026"),
+                WordBox(x0=100, top=0, x1=160, bottom=10, text="Coffee"),
+                WordBox(x0=200, top=0, x1=240, bottom=10, text="$4.50"),
+                # 100 pts away — well past the 25-pt merge ceiling
+                WordBox(x0=0, top=110, x1=200, bottom=120, text="Daily"),
+                WordBox(x0=80, top=110, x1=200, bottom=120, text="Ledger"),
+                WordBox(x0=180, top=110, x1=300, bottom=120, text="Balances"),
+            ]
+            return [Page(
+                page_no=1, width=300, height=200, text="", words=words,
+            )], []
+
+        mod.extract_pages_auto = fake
+        try:
+            rows, _ = scan_pdf_for_transactions(b"")
+        finally:
+            mod.extract_pages_auto = original
+
+        assert len(rows) == 1
+        assert rows[0]["description"] == "Coffee"
+        assert "Ledger" not in rows[0]["description"]
+
+
 class TestMultilineDescription:
     def test_continuation_line_merges(self):
         """A line with no date and no amount, sitting between two