Long-term non-irrigation (N-IRR) and irrigation (IRR) might affect pathogen communities causing seed-rot, root-rot, and damping-off of soybean. Characterization and quantification of Pythium, Fusarium, and Rhizoctonia spp. are critical for understanding the ecology of these microorganisms and management of root diseases. The populations and communities of Pythium, Fusarium, and Rhizoctonia were assessed in soils with long-term N-IRR and IRR using soil dilution plating, denaturing gradient gel electrophoresis (DGGE), and DNA sequence analysis. Disease incidences of seed-rot, root-rot, and damping-off of soybean were also evaluated with the soils from N-IRR and IRR in a growth chamber. Results showed that the population densities of Pythium and Rhizoctonia based on dilution plating and richness based on DGGE were significantly lower in soils with N-IRR than IRR, whereas the population densities of Fusarium and richness of Fusarium were significantly higher in soils with N-IRR than IRR. Cluster analysis based on DGGE band patterns demonstrated that the communities of Pythium, Fusarium, and Rhizoctonia were separated based on N-IRR and IRR. Species diversities of some Pythium, Fusarium, or Rhizoctonia were impacted differently in soils with N-IRR and IRR. Moreover, canonical correspondence analysis (CCA) showed that the disease incidences were positively correlated with levels of soil moisture, pH, the populations of Pythium and Rhizoctonia, richness of Pythium and Rhizoctonia, and negatively correlated with soil porosity, humic matter, and potassium. Long-term N-IRR and IRR could impact pathogen populations and communities and disease incidences differently, and the disease complex might be caused by different species under different systems.