This review aimed to make some considerations about the selection in early generations (seedling generation, first and second clonal generations) and its implications for the potato breeding, which requires the evaluation of thousands of seedlings in order to identify the best genotypes. Considering the high costs of the breeding process, the constant improvement of the selection techniques by the utilization of experimental information is essential. The elimination of the worst genotypes as earlier as possible, by means of selection in early generations, avoids that the inferior genotypes be kept in the population, providing considerable reduction in inputs, field and labor, which means a lower cost. Although there are some studies in the literature reporting low selection efficiency of potato genotypes in early generations, mainly in relation to quantitative characters; other publications show it is efficient. It was concluded that selection in early generations, beginning in the seedling generation, is a viable alternative, allowing higher versatility and efficiency to the potato breeding programs. However, the traits more effective to be selected and the selection pressure applied should be defined in order to avoid the elimination of materials that should be kept in the population.
Solanum tuberosum L; breeding; selection efficiency