Background: Inequity in megacities is a real concern in public health perspective. Tehran is a megacity with more than 8 million population that is divided into 22 regions (counties) with considerable diversity in socioeconomic status. On the other hand, spatial cluster detection is an important tool in disease surveillance for identifying areas of elevated risk and generating hypotheses about disease or mortality etiology. The present research aims to identify high or low-risk clusters for five non-communicable leading causes of death in 22 regions of Tehran province. Methods: Cause-specific mortality rates were extracted from Behesht-e-Zahra registry system for Tehran province in 2011. Spatial scan statistic was chosen as the most common method in spatial cluster detection to detect clusters with elevated risk of death. Given the observed and expected number death in each region, a log likelihood ratio (LLR) criterion was used to test whether a cluster is significant. Result: Two high-risk and two low-risk clusters were detected for each cause of death. All these clusters were statistically significant with P values less than 0.05. Mapping these clusters shows substantial differences between regions in Tehran. For mortality due to ischemic heart diseases, cerebrovascular diseases, hypertensive diseases, respiratory diseases, and stomach cancer, the high-risk clusters concentrated in the southern half of Tehran and low-risk clusters were in the northern half of Tehran. In the most situations, regions 2, 3, and 5 seemed to have lower rates of death compared with other regions. On the other hand, regions, 16, 19, and 20 were in the high rate clusters. Conclusion: There was substantial disparity between regions of Tehran for five non-communicable causes of death studied in this article. Identifying factors affecting the observed differences is useful to set effective preventive interventions and can be investigated in future researches. © 2015 Academy of Medical Sciences of I.R. Iran. All rights reserved.