The effectiveness of one cycle (C) per year of phenotypic recurrent selection was evaluated for improving tolerance of peas (Pisum sativum L.) to aphanomyces root rot (Aphanomyces euteiches Drech.). Each cycle included field-screening a population of F2 lines (summer), diallel intermating among lines selected from the field (fall greenhouse), and one generation of selfing F1 plants (spring greenhouse) to produce F2 lines for the next cycle. The schedule is repeated for each cycle. A blocks-within-replicates design was employed in the field screening of C1 and C2 to improve within-block homogeneity. Selection intensities were 12.4%, 11.1%, and 10.6% for C0, C1, and C2, respectively. Using the performance of a tolerant control line, Mn 108, as a basis of comparison, the realized gain in dry seed yield and survival was 32% and 68% from C0 to C1 and 22% and 115% from C1 to C2, respectively.