A multi-objective evolutionary algorithm with extended minimal generation gap (MGG) model and distance-based density measure is given, which we call DMOEA. DMOEA employs a new technique to estimate the distance between two individuals in the objective space, further, finds the K nearest neighbors on either side of certain individual along one focused objective, calculates the sum of the distances to the K nearest neighbors for the crowding density of the individual, and prunes the set of non-dominated solutions one by one according to the crowding density with the purpose of maintaining the diversity of solutions when the number of non-dominated solutions in the set is more than its capacity. Furthermore, DMOEA extends MGG model with simplex crossover to generate the new population, which can improve the search efficiency. DMOEA has been compared with three other state-of-the-art algorithms, DEMO, NSGA-II, and SPEA2 on a set of representative test problems.