Epidemiological data pertaining to maternal mortality is valuable in each setup to design interventional programs to reduce the ratio favourably. This study was design to evaluate the mortality rate in our hospital, to assess the epidemiological aspects , types of delay, and to suggest recommendations for improvement. Objective of the study was to analyse type of delay, and to suggest measures to reduce it. Methods: A retrospective study done at a tertiary level care centre , data were collected from maternal death review forms and case records. Data studied and analysed. Results: (Majority women were from rural area (73.5.%), and majority (66.8%) were emergency cases .79.3% cases were belonging to lower socioeconomic class and 76.8%cases were uneducated. most cases Type 1 delay was most common (75.4%) comparatively to Type 2 and 3 delay. Conclusions: High risk cases should be identified. Early referral, easy transport, continued skill based training, monitoring of health services can reduce maternal mortality. Special training should be conducted for ASHA workers and ANM who generally works at grass root level in our country. Continued medical training is required for medical officers who are working at PHC and sub district hospital for early recognition of high risk women and their referral in time to higher centers to avoid maternal near miss or death.