Unstack

df.unstack( level=0 )

(The reverse operation is a combination of stack and swaplevel.)