In this talk  I discuss the problem of pointwise density estimation from observations with multiplicative measurement errors. I elucidate the main feature of this problem: the influence of the estimation point on the estimation accuracy. In particular, I show that, depending on whether this point is separated away from zero or not, there are two different regimes in terms of the rates of convergence of the minimax risk. In both regimes I develop kernel–type density estimators and prove upper bounds on their maximal risk over suitable nonparametric classes of densities. It will be shown that the proposed estimators are rate–optimal by establishing matching lower bounds on the minimax risk. Finally  the estimation procedure will be tested on simulated data.