The author critiques two articles by Michael Howlett published recently by this Journal, and comments on policy agenda-setting theory. Howlett's articles use Canadian data to test three different policy agenda-setting theories forwarded by Anthony Downs, by Frank Baumgartner and Bryan Jones and by John Kingdon. This commentary suggests that the tests have several major flaws. These flaws are a product of several factors: Howlett's dataset and use of statistics, his interpretation of the theories at hand and empirical problems with the policy agenda-setting theories themselves.