Abstract: We review the basic, observationally driven requirements for a successful model of particle acceleration in impulsive solar flares and then evaluate the viability of three classes of acceleration mechanisms. We argue that stochastic resonant acceleration is by far the most promising of the mechanisms.