diff --git a/tests/test_accounting.py b/tests/test_accounting.py
index 952a9229..1e22c91d 100644
--- a/tests/test_accounting.py
+++ b/tests/test_accounting.py
@@ -12,12 +12,14 @@ from piker import config
 from piker.accounting import (
     Account,
     calc,
-    Position,
-    TransactionLedger,
-    open_trade_ledger,
+    open_account,
     load_account,
     load_account_from_ledger,
+    open_trade_ledger,
+    Position,
+    TransactionLedger,
 )
+import tractor
 
 
 def test_root_conf_networking_section(
@@ -53,12 +55,17 @@ def test_account_file_default_empty(
 )
 def test_paper_ledger_position_calcs(
     fq_acnt: tuple[str, str],
+    debug_mode: bool,
 ):
     broker: str
     acnt_name: str
     broker, acnt_name = fq_acnt
 
-    accounts_path: Path = config.repodir() / 'tests' / '_inputs'
+    accounts_path: Path = (
+        config.repodir()
+        / 'tests'
+        / '_inputs'  # tests-local-subdir
+    )
 
     ldr: TransactionLedger
     with (
@@ -77,6 +84,7 @@ def test_paper_ledger_position_calcs(
             ledger=ldr,
 
             _fp=accounts_path,
+            debug_mode=debug_mode,
 
         ) as (dfs, ledger),
 
@@ -102,3 +110,87 @@ def test_paper_ledger_position_calcs(
         df = dfs[xrp]
         assert df['cumsize'][-1] == 0
         assert pos.cumsize == 0
+
+
+
+@pytest.mark.parametrize(
+    'fq_acnt',
+    [
+        ('ib', 'algopaper'),
+    ],
+)
+def test_ib_account_with_duplicated_mktids(
+    fq_acnt: tuple[str, str],
+    debug_mode: bool,
+):
+    # ?TODO, once we start symcache-incremental-update-support?
+    # from piker.data import (
+    #     open_symcache,
+    # )
+    #
+    # async def main():
+    #     async with (
+    #         # TODO: do this as part of `open_account()`!?
+    #         open_symcache(
+    #             'ib',
+    #             only_from_memcache=True,
+    #         ) as symcache,
+    #     ):
+
+
+    from piker.brokers.ib.ledger import (
+        tx_sort,
+
+        # ?TODO, once we want to pull lowlevel txns and process them?
+        # norm_trade_records,
+        # update_ledger_from_api_trades,
+    )
+
+    broker: str
+    acnt_id: str = 'algopaper'
+    broker, acnt_id = fq_acnt
+    accounts_def = config.load_accounts([broker])
+    assert accounts_def[f'{broker}.{acnt_id}']
+
+    ledger: TransactionLedger
+    acnt: Account
+    with (
+        tractor.devx.maybe_open_crash_handler(pdb=debug_mode),
+
+        open_trade_ledger(
+            'ib',
+            acnt_id,
+            tx_sort=tx_sort,
+
+            # TODO, eventually incrementally updated for IB..
+            # symcache=symcache,
+            symcache=None,
+            allow_from_sync_code=True,
+
+        ) as ledger,
+
+        open_account(
+            'ib',
+            acnt_id,
+            write_on_exit=True,
+        ) as acnt,
+    ):
+        # per input params
+        symcache = ledger.symcache
+        assert not (
+            symcache.pairs
+            or
+            symcache.pairs
+            or
+            symcache.mktmaps
+        )
+        # re-compute all positions that have changed state.
+        # TODO: likely we should change the API to return the
+        # position updates from `.update_from_ledger()`?
+        active, closed = acnt.dump_active()
+
+        # breakpoint()
+
+        # TODO, (see above imports as well) incremental update from
+        #      (updated) ledger?
+        # -[ ] pull some code from `.ib.broker` content.