Cytologic analysis was performed on 128 bile specimens collected by schedule from 12 liver transplant recipients over a 4-month period. Clinical diagnoses at the time of specimen collection were determined retrospectively, as follows: clinically stable, 75; acute rejection, 15; CMV hepatitis, 1; systemic infection, 8; ischemic injury, 24 (all within the first 4 days postop); nonclassifiable, 5. Bile analysis was done by a blinded investigator. Specimens contained ductal epithelial cells (EC) and inflammatory cells (IC), which were counted using Cytospin slide preparations. Greater than 10 cells/slide were seen in 93.3% of rejections, 91.7% of ischemic injuries, 100% of systemic infections, and 14.6% of stable patients. In samples collected after POD 4, IC were seen in 86.7% of rejections, yielding a specificity of 94.4% (P less than 0.001). If lymphoblastic cells were also seen, the specificity increased to 96.6%. Five specimens were obtained the day before the clinical diagnosis of rejection; all demonstrated IC. Seven specimens were obtained 3 days after beginning therapy for rejection. In 5 the bile contained no IC, and clinical improvement occurred; in the 2 in whom IC were found, further therapy was subsequently required. IC were seen in 5 of 8 specimens taken when systemic infection was present; the clinical setting allowed differentiation from rejection. Only 1 case of CMV hepatitis was included, thus no conclusions can be drawn for this entity. Cytoplasmic vacuolization of EC was observed in 30% of cases, in these, cyclosporine levels were significantly higher (989.9 +/- 356.9 vs. 672.8 +/- 421.2, P = 0.02). In summary, bile cytology analysis aides in the monitoring of the onset and duration of rejection. It may be an indicator of persistent rejection, and it may help prevent overimmunosuppression in those cases with normal cytological findings.