diff --git a/docs/docs/user_guide.md b/docs/docs/user_guide.md
index fbc61b2..29383c7 100644
--- a/docs/docs/user_guide.md
+++ b/docs/docs/user_guide.md
@@ -37,6 +37,9 @@ This method takes 5 optional parameters:
     not available.
     * If set to `False`, `pandarallel` will use multiprocessing data transfer (pipe) to
     transfer data between the main process and workers.
+- `leave`: (bool, `True` by default): whether to keep the progress bar after running on a notebook or not.
+    * If set to `True`, will leave the notebook progress bar after running.
+    * If set to `False`, will delete the notebook progress bar after running.
 
 Using memory file system reduces data transfer time between the main process and
 workers, especially for big data.
diff --git a/pandarallel/core.py b/pandarallel/core.py
index f96359c..5977fb2 100644
--- a/pandarallel/core.py
+++ b/pandarallel/core.py
@@ -205,6 +205,7 @@ def parallelize_with_memory_file_system(
     nb_requested_workers: int,
     data_type: Type[DataType],
     progress_bars_type: ProgressBarsType,
+    leave: bool = True,
 ):
     def closure(
         data: Any,
@@ -239,7 +240,7 @@ def closure(
 
         show_progress_bars = progress_bars_type != ProgressBarsType.No
 
-        progress_bars = get_progress_bars(progresses_length, show_progress_bars)
+        progress_bars = get_progress_bars(progresses_length, show_progress_bars, leave)
         progresses = [0] * nb_workers
         workers_status = [WorkerStatus.Running] * nb_workers
 
@@ -344,6 +345,7 @@ def parallelize_with_pipe(
     nb_requested_workers: int,
     data_type: Type[DataType],
     progress_bars_type: ProgressBarsType,
+    leave: bool = True,
 ):
     def closure(
         data: Any,
@@ -380,7 +382,7 @@ def closure(
 
         show_progress_bars = progress_bars_type != ProgressBarsType.No
 
-        progress_bars = get_progress_bars(progresses_length, show_progress_bars)
+        progress_bars = get_progress_bars(progresses_length, show_progress_bars, leave)
         progresses = [0] * nb_workers
         workers_status = [WorkerStatus.Running] * nb_workers
 
@@ -446,6 +448,7 @@ def initialize(
         progress_bar=False,
         verbose=2,
         use_memory_fs: Optional[bool] = None,
+        leave: Optional[bool] = True,
     ) -> None:
         show_progress_bars = progress_bar
         is_memory_fs_available = Path(MEMORY_FS_ROOT).exists()
@@ -510,36 +513,37 @@ def initialize(
 
         # DataFrame
         pd.DataFrame.parallel_apply = parallelize(
-            nb_workers, DataFrame.Apply, progress_bars_in_user_defined_function
+            nb_workers, DataFrame.Apply, progress_bars_in_user_defined_function, leave
         )
         pd.DataFrame.parallel_applymap = parallelize(
             nb_workers,
             DataFrame.ApplyMap,
             progress_bars_in_user_defined_function_multiply_by_number_of_columns,
+            leave,
         )
 
         # DataFrame GroupBy
         PandaDataFrameGroupBy.parallel_apply = parallelize(
-            nb_workers, DataFrameGroupBy.Apply, progress_bars_in_user_defined_function
+            nb_workers, DataFrameGroupBy.Apply, progress_bars_in_user_defined_function, leave
         )
 
         # Expanding GroupBy
         PandasExpandingGroupby.parallel_apply = parallelize(
-            nb_workers, ExpandingGroupBy.Apply, progress_bars_in_work_function
+            nb_workers, ExpandingGroupBy.Apply, progress_bars_in_work_function, leave
         )
 
         # Rolling GroupBy
         PandasRollingGroupby.parallel_apply = parallelize(
-            nb_workers, RollingGroupBy.Apply, progress_bars_in_work_function
+            nb_workers, RollingGroupBy.Apply, progress_bars_in_work_function, leave
         )
 
         # Series
         pd.Series.parallel_apply = parallelize(
-            nb_workers, Series.Apply, progress_bars_in_user_defined_function
+            nb_workers, Series.Apply, progress_bars_in_user_defined_function, leave
         )
-        pd.Series.parallel_map = parallelize(nb_workers, Series.Map, show_progress_bars)
+        pd.Series.parallel_map = parallelize(nb_workers, Series.Map, show_progress_bars, leave)
 
         # Series Rolling
         pd.core.window.Rolling.parallel_apply = parallelize(
-            nb_workers, SeriesRolling.Apply, progress_bars_in_user_defined_function
+            nb_workers, SeriesRolling.Apply, progress_bars_in_user_defined_function, leave
         )
diff --git a/pandarallel/progress_bars.py b/pandarallel/progress_bars.py
index dfe9190..f3ebe69 100644
--- a/pandarallel/progress_bars.py
+++ b/pandarallel/progress_bars.py
@@ -118,12 +118,13 @@ def update(self, values: List[int]) -> None:
 
 
 class ProgressBarsNotebookLab(ProgressBars):
-    def __init__(self, maxs: List[int], show: bool) -> None:
+    def __init__(self, maxs: List[int], show: bool, leave: bool = True) -> None:
         """Initialization.
         Positional argument:
         maxs - List containing the max value of each progress bar
         """
         self.__show = show
+        self.__leave = leave
 
         if not show:
             return
@@ -159,6 +160,9 @@ def update(self, values: List[int]) -> None:
 
             if value >= bar.max:
                 bar.bar_style = "success"
+                if not self.__leave:
+                    bar.layout.display = 'none'
+                    label.layout.display = 'none'
 
             label.value = "{} / {}".format(value, bar.max)
 
@@ -172,10 +176,10 @@ def set_error(self, index: int) -> None:
 
 
 def get_progress_bars(
-    maxs: List[int], show
+    maxs: List[int], show, leave=True,
 ) -> Union[ProgressBarsNotebookLab, ProgressBarsConsole]:
     return (
-        ProgressBarsNotebookLab(maxs, show)
+        ProgressBarsNotebookLab(maxs, show, leave)
         if is_notebook_lab()
         else ProgressBarsConsole(maxs, show)
     )