Echinococcus granulosus is a cestode species responsible for human cystic echinococcosis (CE). The domestic dog is the primary definitive host for the transmission of E. granulosus infection to domestic livestock intermediate hosts and to humans. In order to determine the prevalence and risk factors for canine echinococcosis, epidemiological studies were conducted in known endemic localities in three different countries i.e. Libya, Kenya and Wales. Stray dogs were subject to direct examination of the small intestine at necropsy. Owned dogs (total n=1659) were screened for E. granulosus infection using a genus specific coproantigen ELISA test that was restandardised against 75 dog necropsy results, and exhibited 100% sensitivity and 98% specificity. Analysis of risk factors for canine echinococcosis associated with Echinococcus coproantigen positive results was evaluated based on dog-owner questionnaires. In Libya, the overall infection of E. granulosus in autopsied dogs was 25.8%, and an overall coproantigen prevalence of 21.6% was obtained for owned dogs (n= 334). Risk factors (p<0.05) for a coproantigen positive owned dog in Libya were, dog age (≤5 years), non-restraint of dogs, dogs fed on offal, sheepdog class, owners that did not de-worm their dogs, and dogs owned by farmers who admitted to frequent slaughter of livestock. Dog sex was not a significant risk factor for a positive coproantigen ELISA in dogs from all study areas. Worm burdens in necropsied dogs in Libya ranged from 29 to 2900 (mean 1064) and were positively correlated to coproantigen ELISA OD values (r= 0.76), but negatively associated with dog age (r= -0.64). Canine echinococcosis in Libya measured by locality varied, with Alkhums (Leptis-Magna) district having the highest coproantigen prevalence at 38.7% (p=0.001), followed by Azahwia district with 19.2%. Tripoli district had the lowest coproprevalence where 17.5% of dogs were copro-positive. In Tripoli an abattoir survey for livestock CE was also undertaken and is reported. In northern Kenya, post-mortem examination of the small intestines of 17 dogs from Turkana, revealed 29.4% harboured E. granulosus infection, with a mean worm burden of 1570 adults. Overall 26% (42/161) of Turkana dogs were copro-positive, with the highest copro-prevalence identified in dogs from Lokichoggio division. Younger dogs (≤5 years), free-roaming dogs, dogs fed on offal, and dogs of owners that practiced home slaughter of livestock and that had no knowledge about echinococcosis, appeared to have a significantly higher risk of being coproantigen positive (p<0.05). In mid-Wales a list of 321 sheep farms were selected at random, from which 1164 farm-dogs were screened using rectal faecal samples tested for Echinococcus coproantigens, and owners questioned using a modified dog-owner questionnaire. Furthermore, the potential impact of the 2001 footand- mouth disease (FMD) epidemic, on the prevalence of E. granulosus in farm dogs was assessed. An overall coproantigen positive rate of 8.0% was recorded on 22% of farms surveyed, which compared to a rate of 3.4% obtained in the same region in 1993. There was no significant difference in copro-positive prevalence between FMD affected properties and those that were unaffected. Significant risk factors for a positive farm dog in Wales were associated with allowing dogs to roam free, and infrequent deworming (>4month intervals) of dogs. The data are discussed in relation to a previous pilot hydatid control program in that area of Wales (1983-89) and the potential for increase in transmission to humans. Identification of risk factors associated with canine echinococcosis appear to be similar in all three communities studied and help to demonstrate practices that may be amenable to change as part of hydatid control programmes.