In many types of weed control experiments most economical progress can be made by visual rating of the plots. While such ratings are admittedly somewhat subjective, this does not necessarily mean that the results are less accurate or less satisfactory than methods which profess to be quantitative. For example, counting thistle sprouts in a square-rod area may show more sprouts in a treated plot than in the check because the herbicide used resulted in a number of weak sprouts replacing every strong sprout killed to the ground line. If only counts are made, the data will look as if no progress was being made, while if both the strength and number of sprouts are rated the actual situation will be much more nearly approximated.