The prevalence of Echinococcus granulosus in Cypriot dogs during the first dog-test period of the anti-echinococcosis campaign (June–December, 1972)_was described and analyzed by contingency table methods and stepwise multiple regression. Mean canine E. granulosus prevalence rates, dogs per person, infected dogs per person, and sheep and goats per dog were found to be higher in Turkish Cypriot villages than in Greek Cypriot villages. It was also observed that during the first dog-test period of the campaign, only one of 285 rural village slaughterhouses was located in a Turkish Cypriot village. Results of a step-wise regression analysis showed that animal husbandry, canine, ethnic and economic variables accounted for 23% of the variation in initial canine E. granulosus prevalence rates. The effects of multicollinearity and measurement error on the results of these analyses and deficiencies in the data were discussed.