Background: Improving perinatal health is key in decreasing child and infant mortality rates but informationrnon perinatal deaths is highly susceptible to omission and misreporting. Hence community based data canrnprovide more representative and reliable information. rnObjective: this Study is intended to assess the magnitude, determinants and main cause of perinatal mortalityrnamong babies born from cohorts of pregnant women in three selected districts of North Shewa zone OromiarnRegional State namely Degem, Kuyu and Warajarso.rnMethods: A cohort of 5,784 pregnant women were followed through delivery by the Maternal and NewbornrnHealth in Ethiopia Partnership (MaNHEP) from March 2011 to February 2012 in three districts of NorthrnShewa zone of Oromia Regional State where MaNHEP is providing maternal and neonatal healthrnintervention. At the beginning of the study, pregnancy outcomes of 3,916 women were identified by house tornhouse visit using the initial records of MaNHEP in each kebele. Using this cohort we conducted communityrnbased nested case-control design among 67 cases and 134 controls. All perinatal deaths reported betweenrnMarch 2011 to February 2012 were included in the study and two live birth controls born from mothers inrnthe cohort were included for each perinatal death after a random selection from the same Gote with therncases. A standard verbal autopsy instrument was used as a main data collection tool after adapting to the localrnsituation. Data were entered using Epi Info version 3.5.1 for windows and exported to SPSS version 20.0 forrnanalysis. Logistic regression model was employed with Crude and Adjusted Odds Ratio at 95 CI, p-value 0.05rntaken as level of statistical significance. The probable causes of deaths were assigned by two pediatriciansrnusing ICD 10. Principal component analysis was used to construct household socio-economic status of thernstudy participants.rnResult: Data were collected from 67 cases and 134 controls. Twenty four (30.38 %) of the cases were stillrnbirths while 43(54.43%) were early neonatal deaths. The overall perinatal mortality rate in the study area wasrn19.5 per 1000 live births and still birth. Maternal age greater than 30 years [AOR 2.665, (95% CI:1.0266.922)],rnrnhaving previous history of neonatal mortality [AOR 6.833, (95% CI:2.782-16.782)], preterm birthsrn[AOR 6.929,( 95% CI:1.779-26.994)], male sex [AOR 4.049,(95% CI:1.836-8.926)] and home delivery [AORrn3.704,(95% CI:1.192-11.506)] were significantly associated with perinatal mortality in the final model. Birthrnasphyxia (31%), sepsis (25%), chorioamnionitis (11%), antepartum hemorrhage (9%) were the four leadingrncauses of perinatal death.rnConclusion: The study reported lower perinatal mortality rate than other studies conducted in the country. rnMajority of the deaths occurred in the first twenty four hours suggesting that interventions should focus onrnthis critical time of vulnerability to save lives of newborn. Potential determinants and causes of perinatalrnmortality identified in this study were related with maternal factors implies that still there is a need to focus onrnthese major causes of deaths for further intervention.