Each Gleason score category of prostatic adenocarcinoma (or Grade Group) may encompass a diverse group of architectural patterns such as well-formed glands, poorly formed glands, cribriform structures, single cells, and/or solid sheets. We have noted heterogeneity within the single-cell subtype of Gleason pattern 5 prostatic adenocarcinoma that has not been fully addressed. Therefore, we retrospectively reviewed a series of radical prostatectomies with high-grade prostatic adenocarcinoma (Grade Group 4 or 5), identifying tumors with a component of single-cell infiltration. Additional cases identified prospectively were also included. TNM status, association with other histologic patterns, and clinical follow-up status were determined. Immunohistochemistry for NKX3.1, E-cadherin, p120 catenin, and prostate-specific antigen (PSA) were performed in each case. Eighteen cases with a component of well-developed Gleason pattern 5 characterized by single infiltrative cells that comprised ≥5% of the tumor were identified (15/202 retrospective radical prostatectomies with the high-grade disease [7.5%]). The single-cell pattern ranged from 5% to 50% of the tumor volume, with 5 cases containing ≥40%, and variable secondary architecture included diffuse infiltrating single cells with targetoid growth pattern around benign glands, solid expansive nests of noncohesive cells, and corded/single file growth pattern. Further morphologic analysis demonstrated 2 distinct histologic subtypes: (1) (subtype 1; n=9) monomorphic “plasmacytoid” tumor cells with eccentrically placed nuclei and variable intracytoplasmic vacuoles with bland cytology and discohesion and (2) (subtype 2; n=9) more cohesive tumor cells with greater cytologic atypia characterized by prominent nucleoli, greater variability in nuclear size/shape, occasional mitotic figures, and more irregular infiltration. By immunohistochemistry, NKX3.1 nuclear expression and PSA cytoplasmic expression was retained in all cases. Concomitant membranous E-cadherin loss and strong cytoplasmic p120 catenin expression were present in 5 of the 18 (28%) cases, all in subtype 1 (5/9, 56%). Overall, 56% (10/18) of patients had advanced-stage disease (≥pT3b), and 70% (7/10) of these patients had associated lymphovascular invasion. All patients had concomitant cribriform patterns of carcinoma. The outcome was available for 14 patients: 4 died of unknown cause; 6 had biochemical recurrence with distant bone metastasis in 5 of the 6; and 4 patients with <3 years of follow-up currently have undetectable serum PSA levels (2 patients received salvage radiotherapy with androgen deprivation and 2 remain on routine follow-up). In summary, the single-cell pattern of Gleason pattern 5 prostatic adenocarcinoma is uniformly associated with other high-risk histologic patterns (eg, cribriform growth), and high-stage disease with distant metastasis is not uncommon. Our data suggest that the “single-cell” Gleason pattern 5 prostatic adenocarcinoma contains 2 distinct subtypes. Somatic CDH1 alterations may play a role in the development of the “plasmacytoid” pattern characterized by monomorphic cytology with concomitant E-cadherin loss and aberrant p120 catenin expression.