diff --git a/freqtrade/data/converter.py b/freqtrade/data/converter.py
index 7f7e79445..98ed15489 100644
--- a/freqtrade/data/converter.py
+++ b/freqtrade/data/converter.py
@@ -119,14 +119,6 @@ def ohlcv_fill_up_missing_data(dataframe: DataFrame, timeframe: str, pair: str)
         else:
             # Don't be verbose if only a small amount is missing
             logger.debug(message)
-    candle_price_gap = 0
-    if not df.empty and 'close' in df.columns:
-        returns = df['close'].pct_change().dropna()
-        if len(returns):
-            candle_price_gap = max(abs(returns))
-    if candle_price_gap > 0.3:
-        logger.info(f"Price jump in {pair} between two candles of {candle_price_gap:.2%} detected.")
-
     return df
 
 
diff --git a/freqtrade/data/history/idatahandler.py b/freqtrade/data/history/idatahandler.py
index 80e29f4c0..19d753043 100644
--- a/freqtrade/data/history/idatahandler.py
+++ b/freqtrade/data/history/idatahandler.py
@@ -335,6 +335,20 @@ class IDataHandler(ABC):
                     "Use `freqtrade download-data` to download the data"
                 )
             return True
+        else:
+            candle_price_gap = 0
+            if (candle_type in (CandleType.SPOT, CandleType.FUTURES) and
+                    not pairdf.empty
+                    and 'close' in pairdf.columns and 'open' in pairdf.columns):
+                # Detect gaps between prior close and open
+                gaps = ((pairdf['open'] - pairdf['close'].shift(1)) / pairdf['close'].shift(1))
+                gaps = gaps.dropna()
+                if len(gaps):
+                    candle_price_gap = max(abs(gaps))
+            if candle_price_gap > 0.1:
+                logger.info(f"Price jump in {pair} between two candles of "
+                            f"{candle_price_gap:.2%} detected.")
+
         return False
 
     def _validate_pairdata(self, pair, pairdata: DataFrame, timeframe: str,
diff --git a/tests/data/test_converter.py b/tests/data/test_converter.py
index 83429bb38..f74383d15 100644
--- a/tests/data/test_converter.py
+++ b/tests/data/test_converter.py
@@ -80,10 +80,10 @@ def test_ohlcv_fill_up_missing_data2(caplog):
     ticks = [
         [
             1511686200000,  # 8:50:00
-            8.794e-07,  # open
-            8.948e-07,  # high
-            8.794e-07,  # low
-            8.88e-07,  # close
+            8.794e-05,  # open
+            8.948e-05,  # high
+            8.794e-05,  # low
+            8.88e-05,  # close
             2255,  # volume (in quote currency)
         ],
         [
@@ -118,7 +118,6 @@ def test_ohlcv_fill_up_missing_data2(caplog):
     assert len(data) == 3
     caplog.set_level(logging.DEBUG)
     data2 = ohlcv_fill_up_missing_data(data, timeframe, "UNITTEST/BTC")
-    assert log_has_re(r"Price jump in .* between two candles .* detected\.", caplog)
     assert len(data2) == 4
     # 3rd candle has been filled
     row = data2.loc[2, :]
diff --git a/tests/data/test_datahandler.py b/tests/data/test_datahandler.py
index 5d6d60f84..4b1f9634c 100644
--- a/tests/data/test_datahandler.py
+++ b/tests/data/test_datahandler.py
@@ -15,7 +15,7 @@ from freqtrade.data.history.idatahandler import IDataHandler, get_datahandler, g
 from freqtrade.data.history.jsondatahandler import JsonDataHandler, JsonGzDataHandler
 from freqtrade.data.history.parquetdatahandler import ParquetDataHandler
 from freqtrade.enums import CandleType, TradingMode
-from tests.conftest import log_has
+from tests.conftest import log_has, log_has_re
 
 
 def test_datahandler_ohlcv_get_pairs(testdatadir):
@@ -154,6 +154,85 @@ def test_jsondatahandler_ohlcv_load(testdatadir, caplog):
     assert df.columns.equals(df1.columns)
 
 
+def test_datahandler__check_empty_df(testdatadir, caplog):
+    dh = JsonDataHandler(testdatadir)
+    expected_text = r"Price jump in UNITTEST/USDT between"
+    df = DataFrame([
+        [
+            1511686200000,  # 8:50:00
+            8.794,  # open
+            8.948,  # high
+            8.794,  # low
+            8.88,  # close
+            2255,  # volume (in quote currency)
+        ],
+        [
+            1511686500000,  # 8:55:00
+            8.88,
+            8.942,
+            8.88,
+            8.893,
+            9911,
+        ],
+        [
+            1511687100000,  # 9:05:00
+            8.891,
+            8.893,
+            8.875,
+            8.877,
+            2251
+        ],
+        [
+            1511687400000,  # 9:10:00
+            8.877,
+            8.883,
+            8.895,
+            8.817,
+            123551
+        ]
+    ], columns=['date', 'open', 'high', 'low', 'close', 'volume'])
+
+    dh._check_empty_df(df, 'UNITTEST/USDT', '1h', CandleType.SPOT, True)
+    assert not log_has_re(expected_text, caplog)
+    df = DataFrame([
+        [
+            1511686200000,  # 8:50:00
+            8.794,  # open
+            8.948,  # high
+            8.794,  # low
+            8.88,  # close
+            2255,  # volume (in quote currency)
+        ],
+        [
+            1511686500000,  # 8:55:00
+            8.88,
+            8.942,
+            8.88,
+            8.893,
+            9911,
+        ],
+        [
+            1511687100000,  # 9:05:00
+            889.1,   # Price jump by several decimals
+            889.3,
+            887.5,
+            887.7,
+            2251
+        ],
+        [
+            1511687400000,  # 9:10:00
+            8.877,
+            8.883,
+            8.895,
+            8.817,
+            123551
+        ]
+    ], columns=['date', 'open', 'high', 'low', 'close', 'volume'])
+
+    dh._check_empty_df(df, 'UNITTEST/USDT', '1h', CandleType.SPOT, True)
+    assert log_has_re(expected_text, caplog)
+
+
 @pytest.mark.parametrize('datahandler', ['feather', 'parquet'])
 def test_datahandler_trades_not_supported(datahandler, testdatadir, ):
     dh = get_datahandler(testdatadir, datahandler)