Dogslife is the first large-scale, longitudinal cohort study of canine lifestyle, morphology and health. The project involves recruiting the owners of UK-based, Kennel Club registered Labrador Retrievers and asking them to submit data about their dogs via an online questionnaire repeatedly as the dogs age. In this thesis, I have analysed Dogslife data regarding the lifestyle, morphology and health of Labrador Retrievers up to four years of age. A validation study was initially undertaken in order to understand the quality of the Dogslife data because this would underlie all future investigations. Owners were visited and veterinary records scrutinised. It was determined that Dogslife illness reports were subject to recall decay and that minor changes would improve the usability of the questionnaire. Accelerometers were subsequently sent to a subset of the cohort and aspects of the Dogslife exercise questionnaire were found to be correlated to accelerometer readings indicative of sedentary, light and moderate to vigorous exercise. Overall, Dogslife dogs were exercised for over two hours each day with the time spent being dominated by time off lead and on other activities. Dogs in England spent less time exercising than those in Scotland and Wales and dogs in family households spent less time exercising than those in single adult households or households comprising more than one adult. Despite being pedigree animals, the males in the cohort were 2-3cm taller than the breed standard. On average, the females met the breed standard but there was wide variation for both sexes. Working dogs in the cohort were over 2kg lighter than household pets and chocolate coloured dogs were 1.4kg heavier than their black and yellow counterparts. Dogs in multi-dog households were 0.5kg lighter than those in households with no other dog. Heavier dogs spent less time fetching, chasing and retrieving and on other exercise. Over 6,000 signs of illness were reported to Dogslife in the first three and half years and approximately half of them did not involve a veterinary visit. Reported signs were dominated by vomiting and diarrhoea, both of which peaked when the dogs were between 3-6 months of age. For the first time, rates of diarrhoea were shown to be positively associated with human population density in the UK. Limber tail was found to be associated with swimming in the cohort and working dogs were more likely to develop the condition than pets. Genetic analyses identified regions of interest that might predispose the dogs to limber tail on chromosomes 6 and 30. Data from the Dogslife project provide a unique resource for investigating the epidemiology of Labrador Retrievers. This thesis creates a platform for all such future investigations.