A very simple and efficient approach to deriving estimates of the convergence rate for the penalty methods is suggested. The approach is based on the application of results of the sensitivity the- ory to optimization problems. The suggested convergence analysis uses either various sufficient optimal- ity conditions without imposing any regularity requirements on the constraints or growth conditions under weakened regularity requirements on the constraints.