1: \begin{abstract}

2: While multilevel Monte Carlo (MLMC) methods for the numerical approximation

3: of partial differential equations with  random  coefficients enjoy great popularity,

4: combinations with spatial adaptivity seem to be rare. We present an adaptive

5: MLMC finite element approach based on deterministic adaptive mesh 

6: refinement for the arising "pathwise"  problems and outline a convergence theory

7: in terms of desired accuracy and required computational cost.

8: Our theoretical and heuristic reasoning together with the efficiency

9: of our new approach are confirmed by numerical experiments.

10: \end{abstract}

11: