Phytophthora root and stem rot has developed in commercial soybean fields since 2006 in Brazil, and cultivars with resistance to this disease have not been targeted for this region. Thus, the Phytophthora sojae pathotypes are expected to have virulence to few if any of the Rps genes. The objectives of this study were to characterize the pathotype diversity of P. sojae in Brazil, determine the distribution of the pathogen and predict which Rps genes will be effective and should be used in breeding programs. Isolates were collected in six states (Rio Grande do Sul, Santa Catarina, Paraná, Mato Grosso do Sul, Minas Gerais, and Goiás). The virulence formulae were based on the response of a differential set with 14 Rps genes (1a, 1b, 1c, 1d, 1k, 2, 3a, 3b, 3c, 4, 5, 6, 7, and 8). None of the 17 pathotypes found was reported previously. The most common virulence formulas were: 1d, 2, 3c, 4, 5, 6, 7 (octal code 05471, representing 24 % of the occurrences); 1d, 2, 3b, 3c, 4, 5, 6, 7 (05671, 13 %); 1b, 1d, 2, 3a, 3c, 4, 5, 6, 7 (25571, 8 %); and 1d, 3a, 5, 7, 8 (01123, 8 %). Percentages of isolates with a susceptible interaction with each Rps gene was Rps1a (3 %), Rps1b (11 %), Rps1c (3 %), Rps1d (100 %), Rps1k (3 %), Rps2 (86 %), Rps3a (32 %), Rps3b (19 %), Rps3c (73 %), Rps4 (70 %), Rps5 (89 %), Rps6 (59 %), Rps7 (100 %), and Rps8 (22 %). There was apparently no relationship between pathotypes and origin. Stacking resistance genes Rps1a, Rps1b, Rps1c, and Rps1k with Rps3b or Rps8 would be highly effective for soybean cultivars targeted for Brazil.