Historical and ecological evidences indicate that maternal care by skilled providers is one of thernkey strategies for maternal survival. However, the rate of maternal service utilizations andrnreduction of maternal mortality are very low in Ethiopia. Several factors, which vary accordingrnto different contexts and operate at different levels, can affect the utilization of skilled maternalrncare. Hence, a level-by-level analysis of all sources of poor skilled maternal care utilization hasrnstrategic relevance for the country.rnObjectivernThis study was designed to describe skilled maternal care utilization and to identify associatedrnfactors operating at different levelsrnMethodsrnThe study included a linked facility and population-based survey to assess the availability,rnutilization and correlates of maternal service utilization, and a prospective cohort study tornevaluate the effect of the available skilled maternal care in reducing adverse pregnancyrnoutcomes. The linked survey was conducted over three months (January - March 2012) in twelvernrandomly selected kebeles, their nearby twelve health centers, and three hospitals of NorthrnGondar Zone, while the cohort study was conducted in 10 kebeles of Dabat district fromrnDecember 1, 2011 to August 31, 2012 on 763 pregnant women. Data on potential correlates ofrnskilled maternal care utilization were gathered using quantitative and qualitative data collectionrntechniques. In the cohort study, data was collected at four time points: first contact, during the 9thrnmonth of pregnancy, within one week after delivery, and on the forty-second day of postpartum.rnDuring analysis, the effect of potential confounders was controlled using logistic regressionrnmodels. The effect of cluster variation and a number of individual, communal (kebele), andrnfacility-related variables for skilled maternal care utilization was examined using the multilevelrnmodelling.rnResultsrnAll indicators of skilled maternal care service utilization were very low in North Gondar. Out ofrnthe total 1668 women, only 32.3%, 13.8%, and 6.3% utilized skilled providers for their routinernXIVrnantenatal, delivery, and postnatal care, respectively. Most of these services were at health centersrnby nurses or midwives. Of the 476 women who faced complications, 248 (52.1%, 95% CI:rn47.6%, 56.6%) sought assistance from a skilled provider.rnAntenatal and delivery care were available in most of the visited facilities. However, importantrncomponents of both the routine and emergency maternity services were incomplete. Signalrnfunctions including the administration of anticonvulsants and assisted vaginal delivery werernmissing in seven and five of the 12 health centers, respectively. The proportion of women accordingrnto the type of service they received at the time of their antenatal care was 79% (blood pressurerncheckup), 35% (urine testing), 45% (tetanus immunization), 64% (iron supplementation), 51%rn(birth preparedness counseling) and 71% (HIV testing). During delivery, only 24% of thernproviders used partograph consistently. In many of the facilities, important pieces of equipmentrnwere either absent or not functional. Most of the health centers also lacked laboratory tests, suchrnas VDRL, hemoglobin, urine protein, and cross match. Interviewed midlevel providers had nornadequate training on essential procedures like assisted vaginal delivery (vacuum extraction),rnmanual vacuum aspiration, or evacuation and curettage. In addition, about one-third of thernproviders had no skill to manage preeclampsia and post-abortion complications.rnThe multilevel analysis showed a significant heterogeneity among clusters for each indicator ofrnskilled maternal care utilization (ANC, delivery and PNC). At the individual level, preference forrnskilled providers and previous experience of antenatal care were consistently strong predictors ofrnall indicators of skilled maternal care utilization. First birth order, maternal education, andrnawareness about health facilities to get skilled professionals were significantly associated withrnskilled antenatal and delivery care utilization. At this level, variables related to awareness andrnperceptions were more important. At communal and facility level; cost requirements at healthrnfacilities, having different sources of income to cover transport and service costs, and thernavailability of important service components at health facilities were among the factors affectingrnskilled maternal care, especially delivery service. Seeking skilled care for complications wasrnsignificantly affected by wealth, ANC use, and the readiness of a woman for possiblerncomplications.rnThe follow-up study showed that the chance of the occurrence of adverse pregnancy outcomesrnwas reduced by 9% (OR=0.91; 95% CI: 0.43, 1.69) among mothers with less than four ANCrnvisits and by 25% (OR=0.75; 95% CI: 0.25, 2.75) among mothers with four or more ANC visits.rnXVrnIn addition, delivery by a skilled attendant showed a 31% (OR=0.69; 95% CI: 0.36, 1.33)rnreduction in the occurrence of complications and death during delivery and postpartum period.rnHowever, the findings were not significant. Providing incomplete service (poor quality of care)rnwas considered as the major explanation for the insignificant effects.rnConclusionsrnMost women did not use skilled care for the routine antenatal, delivery, and postnatal care.rnNearly half of the women who faced complications also did not seek skilled maternal care. Bothrnthe routine and emergency maternity cares lack the full components of important services. Lackrnof facilities in the health institution and deficiencies in the skills of providers were the majorrnreasons for missing essential components of maternal services. In general, the utilization ofrnskilled maternal care depends on the separate and joint effects of individual, communal, andrnhealth facility level factors. Beyond its effect on health-seeking behavior of mothers, providingrnincomplete (poor quality) maternal service was an obstacle to the achievement of the goals ofrnreducing morbidity and mortality. Therefore, the available skilled maternal care showed anrninsignificant reduction in adverse pregnancy outcomes (complications and deaths).rnRecommendationsrnIn order to create better community awareness and perception about skilled providers and theirrncare, safe motherhood education, especially on the risk of pregnancy and the benefits of skilledrnmaternal care should be promoted using the available communication networks in the ruralrncommunities. Such interventions should target underprivileged women. ANC service should bernexpanded through outreach programs and domiciliary service to attain higher coverage in allrntypes of skilled maternal care. Improving the quality of both the routine and emergencyrnmaternity services (providing important service components) is the most urgent intervention tornachieve the goals of maternal health programs. Providing in-service training and the necessaryrnequipment as well as ensuring continuum of care and monitoring the health facility using thernsignal functions in regular bases are the major activities for effective maternal care.