It has been suggested that a consequence of the declining prevalence of smoking in the United States over the past 25 years is that the population of remaining smokers is becoming more difficult to treat. That is, compared to smokers of years past, today's smokers are more likely to be highly nicotine dependent, have comorbid psychiatric and substance abuse disorders, have less education, and be of lower economic status. If this is indeed the case, there should be found declining abstinence rates in published clinical trials, when the type of treatment is held constant. To test this hypothesis, a literature search was conducted for group-administered multicomponent smoking cessation trials that emphasized coping skills training. Twenty-three studies published between 1977 and 1996 were identified. As predicted, robust negative correlations were found between year of publication and end-of-treatment abstinence rates. Using point-prevalence abstinence rates from later follow-up points produced somewhat weaker associations. Controlling for the use of biochemical verification or nicotine replacement therapies did not alter the findings. In summary, the efficacy of clinical trials--with treatment held constant--appears to be declining. One possible cause is the increasing recalcitrance of those individuals who continue to smoke despite social, regulatory, and medical pressures to quit.