A retrospective 7-mo study (June 1, 1993-December 31, 1993) was performed to evaluate the significance of reporting atypical squamous cells of undetermined significance (ASCUS) on cervical smears. During this period 114 (0.7%) of 17,330 smears were diagnosed as ASCUS, 62 of which resulted in either follow-up cervical biopsies and/or endocervical curettages (61%) or repeat smears (39%), and form the basis of this study. Tissue specimens (38 cases) were negative in 24 cases (63%) and were positive in 14 cases (37%). Of the positive specimens, there were 9 low grade squamous intraepithelial lesions (LSIL) and 5 high grade squamous intraepithelial lesions (HSIL). Repeat smears (24 cases) were negative in 19 cases (79%) and were persistent for a lesion in 5 cases (21%). The persistent lesions consisted of 2 ASCUS and 3 LSIL. Overall, of the 62 cases, 43 (69%) were negative and 19 (31%) showed a lesion on follow-up tissue specimens or smears. Of the 62 cases, ASCUS diagnoses included cellular changes suggestive of human papillomavirus infection in 17 (27%) cases, nondiagnostic cellular changes in 16 (26%) cases, atypical squamous metaplasia vs. a squamous intraepithelial lesion (SIL) in 15 (24%) cases, inflammatory cell changes vs. a LSIL in 5 (8%) cases, 2 of which contained candida, and atrophic cellular changes in 1 (2%) case. In the remaining 8 (13%) cases, proper cytologic analysis was hampered by partially obscuring acute inflammation (3 cases) and low cellularity (5 cases). The results of this study would indicate that the reporting of ASCUS on cervical smears does lead to the detection of SILs in a significant number of cases (31% in this study). Therefore, further evaluation of the patient is suggested.