df_grouped = df.groupby('column') # iterate over each group for group_name, df_group in df_grouped: # iterate over each row for row_index, row in df_group.iterrows(): # do your stuff