df = df.drop_duplicates('COL2') #same as #df = df.drop_duplicates('COL2', keep='first') print (df) COL1 COL2 0 a.com 22 1 b.com 45 2 c.com 34 4 f.com 56