1: \begin{abstract}

2: \noindent We  consider  a  one-dimensional   recurrent  random  walk  in  random

3: environment (RWRE)  when the environment is i.i.d.  with a parametric,

4: finitely supported distribution. Based  on a single observation of the

5: path,  we provide  a maximum  likelihood estimation  procedure  of the

6: parameters of the environment. 

7: Unlike most of the classical maximum likelihood approach, the limit of

8: the criterion  function is in general a  nondegenerate random variable

9: and convergence does not hold in probability.  Not only

10: the leading  term but also the  second order asymptotics  is needed to

11: fully identify the unknown  parameter. We present different frameworks

12: to illustrate  these facts. We also explore  the numerical performance

13: of our estimation procedure.

14: \end{abstract}

15: