Should I check an ICommand's CanExecute method before calling Execute from procedural code?

When using ICommands in XAML, WPF uses the CanExecute method to enable or disable controls associated with the command. But what if I am calling Execute from procedural code? Should I first check CanExecute to make sure that the command can execute, or should Execute take care of this check for me?

In other words, should I do this:

if (someCommand.CanExecute(parameter, target))
    someCommand.Execute(parameter, target);

Or just this:

someCommand.Execute(parameter, target);