In this paper we contrast a number of univariate models of Canadian GDP. Ou
r preferred models are used to provide a business cycle chronology for Cana
da, which is compared with some existing, more judgmentally determined chro
nologies. We find that a simple, 'two quarters of negative growth' rule for
determining recession dates is the mast similar to our chronology. We also
find that the mast recent recession in Canada was unique in both its lengt
h and the slow speed of recovery. JEL Classification: C22, C51, C52, E32.