Spontaneous reporting remains the most used and, undoubtedly, the most cost-effective approach for the identification of adverse drug reactions (ADRs). Most of the limitations of this method are well recognised but the possibility of receiving false-positive reports of coincidental drug-event associations has received little attention. In this paper we propose a method based on the Poisson distribution for computing the maximum number of reports of an ADR that could be expected to be reported coincidentally. Three parameters are required: (i) the background risk of the event in the reference population, (ii) the total number of patients treated with the drug considered and, (iii) the proportion of cases that have been reported to the pharmacovigilance system. For most empirical situations occurring in the post-marketing surveillance setting, the expected number remains low and only a maximum of one to three cases could be accepted as possibly coincidental. For rare adverse events such as agranulocytosis or toxic epidermal necrolysis, coincidental associations are so unlikely that a number of reports greater than three constitutes a strong warning and requires further investigation. These findings suggest that for rare events, reports of coincidental drug-event associations are too unlikely to be considered as an important limitation of spontaneous reporting.