The most important objective of lentil breeding programs is to develop new genotypes that are genetically more productive. Besides, it is necessary that the varieties obtained have short flowering cycles to allow the later sowing of summer crops. Selection is based through phenotypic means; however, we argue it should be based on genetic or breeding values because quantitative traits are often influenced by environments and genotype–environment interactions. The objectives of this study were to: (i) identify genotypes with the highest merit; (ii) estimate genetic parameters to know the genetic control of morphological traits in macrosperma and microsperma lentil types using best linear unbiased prediction (BLUP). Twenty-five recombinant inbred lines (RILs) from six F4 families selected on the basis of precocity and high yields were tested in four environments for important quantitative traits. The analysis of variance showed significant differences between genotypes, environments, and genotype–environment interactions for all the traits. Seven macrosperma- and two microsperma-type RILs were selected. Based on average ranking from breeding values and molecular data obtained with sequence-related amplified polymorphism (SRAP), the same genotypes were selected. Genotypic coefficients of variation, heritability across and by environment, and genetic correlation coefficients using BLUP were obtained. According to our results BLUP could replace molecular analysis methods because the selection process was simpler, more cost-effective, and more accurate. The breeding value of parents would give a better ranking of their genetic value than would their phenotypic value; therefore, the selection efficiency would be enhanced and the genetic gain would be more predictable. The selected genotypes could become potential commercial varieties or be used as parental lines in future hybridization programs.