Coffee drinking and cigarette smoking are strongly correlated behaviors which have been suggested to act synergistically to produce adverse health consequences, particularly coronary heart disease (CHD). We studied in smokers the influence of four days of multiple daily doses of coffee containing different doses or no caffeine on cigarette smoking behavior, nicotine intake from smoking, heart rate and blood pressure, circadian serum glucose, and urinary catecholamine excretion. We observed a tendency toward greater cigarette consumption during caffeine consumption, and a tendency toward higher plasma nicotine levels during low-dose caffeine compared with the no-caffeine condition; however, these effects were small. No caffeine effects on any other of the above parameters were observed. Previously published research has usually studied effects of single doses of caffeine, which does not account for development of tolerance to effects of caffeine. If caffeine does contribute to CHD risk, it is not likely to be related to caffeine effects on smoking behavior, nicotine intake, blood pressure, heart rate, glucose tolerance, or catecholamine release. Adverse effects of long-term caffeine consumption on lipids cannot be excluded.