An attempt was made to reduce the cigarette smoking of three subjects by means of a special cigarette case that delivered aversive shock when opened. The number of cigarettes smoked was recorded by a counter in the cigarette case. The validity of the counter readings as a measure of smoking was obtained by a specially designed participant‐observer technique. It was found that the rate of smoking decreased as a function of the intensity of the shock. Also, the smoking returned to its previously unpunished level after the shock punisher was discontinued. Both of these findings confirm the results of laboratory studies of punishment of simpler responses and extends them to more complex responses in a naturalistic situation. Surprisingly, the duration for which the apparatus was worn also decreased as a function of the intensity of the shock. This finding reveals that this aversive shock technique produced avoidance behavior that prevents the technique from having extensive applicability for eliminating smoking. The same limitation may apply to the use of aversive shock for eliminating other undesirable behavior