In Canada economies of scale in credit unions come not only from large single office arrangements but from external economies realized from belonging to central provincial credit unions. Making use of aggregate provincial time series and cross-sectional data, this study begins by employing a sequential Akaike's Information Criterion (AIC) test to select the most appropriate model. This procedure permits the isolation of economies of scale from technological change effects. For all provinces, economies of scale are discovered to be significantly different from 1, and for five of the eight provinces examined, technological change was statistically significant from zero. The larger the provincial organization, as illustrated by the Quebec caisses populaires, the higher we find estimates of returns to scale and technological change. An implication may be that both expansion and more centralization should be encouraged and that other provinces may be able to increase efficiency by imitating some of Quebec's operational and administrative practices.