Enhanced knowledge on longevity and mortality in cats should support improved breeding, husbandry, clinical care and disease prevention strategies. The VetCompass research database of primary-care veterinary practice data offers an extensive resource of clinical health information on companion animals in the UK. This study aimed to characterise longevity and mortality in cats and to identify important demographic risk factors for compromised longevity. Crossbred cats were hypothesised to live longer than purebred cats. 3 Descriptive statistics were used to characterise the deceased cats. Multivariable linear regression methods investigated risk factor association with longevity in cats that died at or after 5 years of age. From 118,016 cats attending 90 practices in England, 4,009 cats with confirmed deaths were randomly selected for detailed study. Demographic ch haracterisation showed that 3,660 (91.7%) were crossbred, 2,009 (50.7%) were female and 2,599 (64.8%) were neutered. The most frequently attributed causes of mortality in cats of all ages were trauma (12.2%), renal disorder (12.1%), non-specific illness (11.2%), neoplasia (10.8%) and mass lesion disorders (10.2%). Overall, the median longevity was 14.0 years (IQR 9.0-17.0; range 0.0-26.7). Crossbred cats had a higher median longevity than purebred cats (median (IQR) 14.0 years (9.1-17.0) versus 12.5 years (6.1-16.4), P < 0.001) but individual purebred cat breeds varied substantially in longevity. In cats dying at or after 5 years (n = 3,360), being crossbred, lower bodyweight, neutered and non-insured were associated with increased longevity. This study described longevity in cats and identified important causes of mortality and breed-related associations with compromised longevity. Key words feline, longevity, electronic patient record, mortality, cat, breed, demography, purebred, crossbred, veterinary 4 Introduction Improved understanding of longevity and mortality in pet cats will support enhanced breeding, husbandry and clinical strategies that advance the health and welfare of owned cats (1-3). There are an estimated 8.5 to 10.3 million owned cats in the UK, with 19.0-25.5% of households owning a cat (4, 5), suggesting substantial population impacts from even moderate welfare gains. Recently, health issues associated with purebred dog breeding have been highlighted (6-8). However, breed-related disorder predispositions also affect cats (9-11) and warrant exploration of effects on longevity and mortality (12, 13). Hybrid vigour describes superior viability, production and fecundity of crossbred progeny compared with their purebred parents and is an accepted phenomenon in production species (Dechow et al., 2007; Nicholas, 2010). A recent report that hybrid vigour may influence longevity in dogs (14) indicates that cats may similarly be affected. Despite being stated to have a long lifespan (15), few peer-reviewed reports have described population longevity values and mortality in domestic cats (2). Analysis of veterinary clinical records from 460,000 cats in the US identified an average longevity of 12.1 years and that neutering was associated with extended longevity (16). Analyses of Swedish pet insurance records relating to 49,450 cats identified mortality variation between breeds but not between the sexes (17). However, differing population and 5 study design characteristics thwart subsequent attempts at generalisation to support disorder prioritisation and strategies to minimise disorder impacts (2). Epidemiological analysis of electronic patient records (EPRs) collected from a large sample of primary-care veterinary practices into a single national surveillance system has been recommended to investigate companion animal health (6, 18). Veterinary EPRs provide longitudinal collection of clinical data that are contemporaneously recorded by veterinary health professionals and cover all presented patients and disorders (18). The VetCompass (http://www.rvc.ac.uk/VetCompass/) database of merged primary-care practice EPRs holds an extensive resource of clinical health information on companion animals in the UK (19) and has been interrogated to report on longevity and mortality in dogs (14) . Using the VetCompass database, this study aimed to estimate longevity, report common causes of mortality and identify demographic risk factors for compromised longevity in cats in England. Longevity in crossbred cats was hypothesised to exceed purebred cats. Materials and methods The VetCompass project collects de-identified EPR data from primary-care veterinary practices in the UK for companion animal health surveillance (20). The current study included all cats with data uploaded to VetCompass from September 01, 2009 to December 20, 2012. Practice selection required willingness to participate and use of an appropriately configured practice management system (PMS). Clinicians selected 6 summary diagnosis terms at episodes of clinical care from an embedded VeNom Code list (21). Clinical data extraction from PMSs used integrated clinical queries (22) before uploading to a secure structured query