This thesis aimed to analyze the risk factors associated with the frequency of suspected deaths from urban arboviruses - Dengue, Zika and Chikungunya, in the state of Pernambuco during the epidemic and post-epidemic period - 2015 to 2018. An epidemiological, cross-sectional and descriptive study was carried out on the deaths confirmed by the state committee for discussing deaths from arboviruses in the 12 development regions of the State of Pernambuco. Data (cases and deaths by region; age group, sex, period of clinical evolution; presence of comorbidities; symptoms before death, therapeutic itinerary. diagnosis and entomological monitoring) were collected secondarily from the database of the epidemiology sector of the Arboviruses Management of the Pernambuco State Health Department. In the period, Pernambuco presented a frequency of 239,956 cases of arboviruses, with 183,078 cases of dengue, 56,769 of chikungunya and 109 of zika. 726 deaths were reported and 225 confirmed, 88 (39.1%) due to dengue; 126 (56.0%) had chikungunya, 11 (4.9%) had co-infection with dengue and chikungunya, with no confirmed deaths from zika. The deaths occurred in 10 of the 12 development regions, with no confirmation in Sertão do Araripe and Sertão do São Francisco. The Metropolitan Region had the highest occurrence with 145 deaths (64.5%) and the highest lethality (0.13%). The Mata Norte region had the highest overall mortality rate (3.96/100,000 inhabitants). Deaths from coinfection were distributed in the Metropolitan Region (N: 09/ 82.0%); in the Northern Agreste Region (N: 01/ 9%) and in the Central Agreste (N: 01/ 9.0%). and 121 (53.8%) of the patients who died were male and 104 (46.2%) were female. 34.0% (N: 76) were between 60 and 79 years old. 74.0% (N: 166) had some type of comorbidity, with arterial hypertension being the most frequent (N: 48 / 29.0%). The period of clinical evolution of the disease to the occurrence of death ranged from 0 - 168 days, with an average of 34.7 days, in relation to the total number of deaths (N: 225). Deaths confirmed by dengue had a clinical course of 0 - 48 days, with a mean of 12.3 days; by chikungunya from 0 – 168 days, with a mean of 34.5 days and by co-infection from 0 – 34 days, with a mean of 15.2 days. Among the symptoms presented by patients during the clinical course before death, fever was the most frequent, reported in 86.2% (N: 194), followed by myalgia (N: 137/60.8%) and arthralgia. (N: 133/59.1%). Bleeding symptoms were reported in 32.8% (N: 74) and were reported in 47.2% (N: 42) of dengue deaths, 23.8% (N: 30) of chikungunya deaths and 18.1% (N: 02) of deaths from coinfection. Neurological symptoms were reported in 44.4% (N: 100) of confirmed deaths, distributed in 34.0% (N: 30) of dengue deaths; 50.7% (N: 64) of chikungunya deaths and 54.5% (N: 06) of deaths confirmed with co-infection. Regarding the therapeutic itinerary followed by the patients during the clinical evolution of the diseases, 23.6% (N: 53) were assisted with only 01 hospital care before the occurrence of death, 2.6% (N: 06) were assisted in some At the moment by Primary Care, 35.5% were hospitalized and 2.2% (N: 05) of the deaths occurred at home. 47.0% (N: 106) of the total number of deaths were reported exclusively through the Death Verification Service (SVO), 7.7% (N: 17) were reported by the Death Occurrence Unit (UOO) and the SVO; 36.0% (N: 81) exclusively by the Death Occurrence Unit and 9.5% (N: 21) by the Municipal Health Surveillance. It was found that 82.2% (N: 185) of the deaths had a laboratory diagnosis, and that 56.75% (105) were requested by the SVO and immunohistochemistry was the evaluation method with a reagent result for the most frequent arboviruses (N: 55/29.7%). In 2015 and 2016, dengue serotypes were isolated: DENV-1 and DENV-3, in 2017 DENV-2 was isolated, with no record of isolation in 2018. Entomological monitoring did not reveal any correlation between the rate of Aedes aegypti building infestation and the occurrence of arboviruses, not rejecting the null hypothesis. The data allow us to conclude that the understanding of the epidemiological profile of the occurrence of deaths in epidemics with simultaneous circulation of arbovirus is of great relevance for the qualification of health services, both in the care of affected patients and in the prevention and vector control actions.