Using the rate per 100,000 total population of those actually in custody serving sentences, the 1995 imprisonment rates across the 10 Provinces varied from a low of 76 per 100,000 (for Ontario) to a rate more than double that -- 193 per 100,000 -- for Saskatchewan. An additional complexity is that those Provinces with similar overall incarceration rates (Quebec, Ontario, and British Columbia) ship a significantly different proportion of their imprisoned offenders to be housed in federally supported penitentiaries. Quebec, for example, houses only approximately 36 percent of its contribution to the prison population; whereas, Ontario and British Columbia house 55 percent and 63 percent (respectively) of their prisoners provincially. The rate of custody (provincial and Federal) on any given day is not significantly related to that Province's combined Federal and provincial prison admission rate. Instead, the number of long-term admissions to penitentiaries (sentences of over 2 years) correlates highly with combined Federal and provincial prison counts. Thus, an attack on long-term sentences is more likely to result in sizable reductions in the number of people in custody than will the substitution of non-prison sanctions for short prison sentences. The differences in imprisonment rates among the Provinces do no disappear when expressed in rates per 1,000 reported crimes or people charged, and the relationships among various measures of crime and prison counts are low and inconsistent. 4 tables, 8 notes, and 26 references