Following a brief examination of the main theories of government growth Wagner's law emerged as the dominant theory and tested in the Canadian context, where both exhaustive and transfer outlays and their major subcategories were covered. Based on the statistical properties of the series the tests were carried out in terms of growth rates. Limited support for Wagner's law in four of its major versions for the insignificant category of transfers to business, and in two versions for exhaustive expenditures alone and in combination with total transfers, was restricted to 1926-1988. No support was established for 1947-1988, the period of real government growth. This absence of statistical robustness undermines confidence even in these cases where some support for this hypothesis was found.