Ki67 scoring is required for the grading of pancreatic neuroendocrine tumors. Ongoing debate exists about the best scoring method in terms of accuracy and practicality. Manual counting of cells in camera-captured/printed images is a widely used and accepted method and considered the most reliable one among the manual methods. It requires counting 500 to 2000 cells to determine the Ki67 score accurately and it is time and energy consuming. We investigated the possibility of achieving the same results by counting only a particular fraction of tumor cells in a printed image in a series of 45 (24 grade 1 and 21 grade 2) pancreatic neuroendocrine tumors. After counting Ki67-positive tumor cells in the whole image, the tumor cells were counted within one-tenth of the same image with the aid of a previously prepared grid on an acetate sheet. The cell number obtained was multiplied by 10 to estimate the total cell count and the Ki67 score was calculated. The agreement between the results of the acetate grid and conventional whole-image counting method was assessed. Near-perfect agreement was achieved regarding the total cell count and Ki67 score. The agreement on tumor grade between the two methods was perfect. The time spent on the process was significantly less than that spent on the conventional method. Although it needs to be validated in a larger series, the acetate grid method might be considered an alternative method for Ki67 scoring in neuroendocrine tumors.