RÉSUMÉ
The present study was a cross-sectional type of descriptive one carried out with the objective of determining the diagnostic accuracy of imprint cytology and frozen section of central nervous system tumors. A purposively selected sample of 33 clinically diagnosed patients with CNS tumors has been included for the study. The relevant data on CNS tumors were collected by using imprint cytology, frozen section and paraffin section. The sensitivity of imprint cytology was 75.0% and specificity was 55.17% respectively. On the other hand positive predictive value of imprint cytology was 18.75% and negative predictive value was 94.11%. The sensitivity of frozen section was 100% and specificity was 86.21% respectively. Comparing the findings of the frozen section to histopathology, the positive predictive value was 50.0% and negative predictive value was 100%. The sensitivity and specificity of imprint cytology was lower in comparison to frozen section and paraffin section. Still then imprint cytology is preferred, because it can be carried out rapidly and easily. It will also be a very helpful aid especially when facilities for frozen section are limited but neurosurgical-procedure is available. A well designed research with adequate sample size should be carried out to get better diagnostic accuracy of imprint cytology in central nervous system tumors.