Pig production from Iberian type has increased in last decade as far as about 1.5 x 10⁶ annual slaughtered animals. It is mainly based in crossbred animals using breeds with better productive performance (Duroc), though several production systems are involved. This situation restricts the objectives and methods for its possible genetic improvement, which must be linked to herds under extensive systems with long cycles and final products adressed to the market segment of greatest quality. Within this context, owing to the remarkable genetic differences between Iberian strains described in this review, substitution and crosses between lines are proposed as the initial method of improvement, to avoid the risk of loss of rusticity and fat and meat quality. Selection for lean growth has proved to be effective in an experimental assay, but the improvement through selection requires a rigorous evaluation as, due to the singularity of this population, it corresponds to a situation of maximum aversion to risk. To reduce it, high priced programmes are required: large effective number of breeding animals and/or records of quality parameters, as well as the development of structures combining selection and genetic diffusion.