Abstract Rational The ability of conditioned stimuli to affect instrumental responding is a robust finding from animal as well as human research and is assumed as a key factor regarding the development and maintenance of addictive behaviour. Objectives While it is well known that stress is an important factor for relapse after treatment, little is known about the impact of stress on conditioned substance-associated stimuli and their influence on instrumental responding. Methods We administered in the present study a Pavlovian-to-instrumental transfer (PIT) paradigm with stimuli associated with smoking- and chocolate-related rewards using points in a token economy to light to moderate smokers who also indicated to like eating chocolate. After completion of the first two phases of the PIT paradigm (i.e. Pavlovian training and instrumental trainings), participants were randomly allocated to the socially evaluated cold pressor test or a control condition before the final phase of the PIT paradigm, the transfer phase, was administered. Results The presentation of a smoking-related stimulus enhanced instrumental responding for a smoking-related reward (i.e. ‘smoking-PIT’ effect) and presentation of a chocolate-related stimulus for a chocolate-related reward (i.e. ‘chocolate-PIT’ effect) in participants aware of the experimental contingencies as indicated by expectancy ratings. However, acute stress did not change (i.e. neither enhanced nor attenuated) the ‘smoking-PIT’ effect or the ‘chocolate-PIT’ effect, and no overall effect of acute stress on tobacco choice was observed in aware participants. Conclusions The established role of stress in addiction appears not to be driven by an augmenting effect on the ability of drug stimuli to promote drug-seeking.