OBJECTIVE: To evaluate AgNOR size and dispersion as alternate methods to AgNOR counts in order to differentiate malignant from non-malignant effusions. DESIGN: Comparative study. PLACE AND DURATION OF STUDY: Department of Pathology, Postgraduate Medical Institute, Lahore, from January 2003 to June 2004. MATERIALS AND METHODS: A total of 240 samples of pleural and peritoneal effusions were centrifuged, deposits smeared on slides and stained with H and E and AgNOR stain. The diagnosis of malignancy or otherwise was made on H and E staining. AgNOR counts, variation in size and dispersion of AgNOR dots in smears were graded and compared in malignant and non-malignant effusions. RESULTS: Mean AgNOR counts of 11.47+/-3.60 and 11.04+/-3.89 in malignant pleural and peritoneal effusions, respectively, were significantly (p<0.0001) greater as compared with counts of 3.36+/-0.69 and 3.35+/-0.66 in non-malignant effusions. AgNOR size and dispersion were of higher grade in significantly greater proportion of malignant as compared with non-malignant effusions (p<0.0001). CONCLUSION: Typing of AgNOR size and dispersion was found to be an easy and reproducible alternative to traditional AgNOR counts for differentiating malignant from non-malignant effusions. These parameters should be correlated with already established but expensive techniques of AgNOR area and size imaging by electron microscopy and flow cytometry, as an economical alternative.