df['new_col'] = df.A.apply(lambda x: x[0]) print df A new_col 0 [1, 2] 1 1 [3, 4] 3 2 [8, 9] 8 3 [2, 6] 2