In [28]: df.groupby(df.columns.tolist(),as_index=False).size() Out[28]: one three two False False True 1 True False False 2 True True 1 dtype: int64