We study optimal stopping with exponentially distributed exercise lag. We formalize the problem first in a general Markovian setting and derive a set of conditions under which the solution exists. In particular, no semicontinuity assumptions of the payoff function are needed. We analyze also...