] Chapter 10 
Sensitivity of the Federal Trade Commission 
Test ^4ethod to .Analytical Parameters 
' Michael R. Guerin 
I INTRODUCTION The Federal Trade Commission (FTC) test method for determining 
j the tar, nicotine, and carbon monoxide yields of commercial cigarettes 
I was designed to characterize and compare brands. Relevance to human 
smoking was a consideration in choosing the test method, but the principal 
objective was to select a method that provided the most accurate and 
reproducible result. Relevance to human smoking was addressed by using 
intermittent puffing and by choosing puff volume, puff duration, puff 
frequency, and butt length based on observations of human smokers. 
Accuracy and reproducibility were addressed by selecting a single set of 
smoking conditions, demanding narrow tolerances for variation in the 
conditions, and standardizing everything from cigarette selection, to the 
smoking environment, to the laboratory analytical chemical methods. 
Requirements associated with producing a standard method tend to 
conflict with those associated with maximizing relevance to the human 
situation. Bradford and colleagues (1936) recognized from the beginning 
that humans smoke cigarettes in different and varying ways, but a 
standardized procedure requires that variables be set and controlled. 
For practical purposes, only one set of conditions could be selected. 
At least two factors have led to an increased concern about the relevance 
of the FTC test procedure. First, FTC results increasingly have been viewed 
as a measure of human exposure and therefore health risk. The problem 
is compounded by the assumption that even a small difference in FTC 
results signifies a meaningful difference in human exposure. Second, a much 
greater variety of cigarettes is available today. They range from nonfilter and 
filter cigarettes similar to those available when the method was adopted, to 
j increasingly popular products with very low FTC yields. Behavioral research 
j has demonstrated that low-yield products are consistently smoked differently 
than are higher yield products (Kozlowski et al., 1989). 
This chapter reviews the nature of the FTC test procedure and the 
influence of changes in its specifications on yields. Smoking parameters 
likely to be different for humans from FTC machine smoking are emphasized. 
STANDARD The quantities of tar, nicotine, carbon monoxide, and other constituents 
i MACHINE in cigarette smoke are measured using smoking machines. One or more 
j SMOKING cigarettes are smoked by a machine, the constituents of interest are collected 
' in a suitable trap, and the contents of the trap are chemically analyzed, 
j The quantity in the trap is divided by the number of cigarettes smoked to 
j compute a yield (or delivery) per cigarette. In the case of the FTC procedure. 
