Acinar atrophy and postatrophic hyperplasia in the prostate are commonly confused with adenocarcinoma. The converse situation may also present a diagnostic dilemma. We recently encountered a number of cases of adenocarcinoma with features that mimicked atrophy, raising the serious concern for the underdiagnosis of malignancy. To investigate the frequency of prostatic adenocarcinoma with atrophic features and the histologic criteria that allow its distinction from benign processes, we reviewed the histopathologic findings in 202 consecutive totally embedded whole-mount radical prostatectomy specimens with adenocarcinoma, 100 consecutive routine needle biopsy specimens, and five additional selected needle biopsy specimens. None of the patients had received androgen deprivation therapy before specimen acquisition. Prostatic adenocarcinoma with atrophic features was defined as a proliferation of malignant acini that architecturally resembled atrophy or postatrophic hyperplasia but retained the diagnostic cytologic features of cancer. The acini were round, often dilated and distorted, and lined by flattened attenuated epithelium with scant cytoplasm. All cases had cytologic evidence of malignancy, including nuclear enlargement and prominent nucleoli; these findings could not be attributed to inflammation or treatment effect. Atrophic features were identified in cancer in six radical prostatectomy specimens (3%) and two routine needle biopsy specimens (2%). The proportion of cancer with atrophic features comprised a mean of 27% of each tumor in the prostatectomy specimens (range 10-60%) and 24% in the needle biopsies (range 10-90%). In the prostatectomy cases, the Gleason score of the cancers was 7 (in five cases) and 5 (in one case); in the biopsy specimens the Gleason score was 6 (in five cases) and 7 (in two cases). In addition, atrophic cancer in the prostatectomy cases had luminal eosinophilic proteinaceous secretions (six cases), blue mucin (five cases), crystalloids (two cases), apocrine blebs (three cases), collagenous micronodules (one case), and high-grade prostatic intraepithelial neoplasia within two high-power fields (three cases); the histologic features were similar in the needle biopsy specimens. We conclude that prostatic adenocarcinoma with atrophic features is an unusual finding that is easily confused with benign acinar atrophy. It is recognized by a combination of architectural and cytologic findings and usually coexists with typical Gleason score 5-7 acinar adenocarcinoma. This pattern is important to recognize to avoid the underdiagnosis of malignancy.