In this review we present a brief discussion on the observational evidence in favor of the presence of temperature variations, and conclude that many planetary nebulae show spatial temperature variations that are larger than those predicted by 1D static chemically homogeneous photoionization models. To determine accurate chemical abundances it is necessary to know the cause of these temperature variations and several possibilities are discussed. The importance of this problem is paramount to test the models of stellar evolution of low and intermediate mass stars and of the chemical evolution of galaxies. We conclude that the proper abundances for chemically homogeneous PNe are those derived from recombination lines, while for the two-abundance nebular model the proper heavy element abundances relative to hydrogen are those derived from visual and UV collisionally excited lines adopting the $t^2$ values derived from $T_e$([O III]) and $T_e$(Balmer).