Product recalls have become an almost familiar phenomenon in a consumer's daily life. The reasons for this include the increasingly advanced and at the same time vulnerable technology of consumer products and the emergent law of strict liability of manufacturers and dealers in the US and Europe. This paper deals with the seller's choice situation. The seller has more or less vague evidence of potential damages which could possibly be related to the use or misuse of his products, altogether indicators of a potential product-related crisis. In such a situation, the seller must decide whether to initiate an immediate product recall or to continue with business as usual, whereas the second option may involve a recall at a later date. The analysis of the choice situation is a preliminary one. The academic literature concerning recall behavior is rather sparse. The first part of the paper, therefore, merely collects and discusses some of the relevant factors pro and contra product recalls, as mentioned in various reports and case analyses. The result is a system of factors, such as evidence of damage and of a liability risk, efficiency and speed of a recall, and the availability of other measures, e.g. warnings, the relative importance of the product (substitutes in sight?), significant and minor influence of relevant media, the seller's reputation and ability, the willingness of consumers to cooperate, etc. The second part of the paper takes a closer look at two of the previously mentioned factors: the seller's reputation and the time span between gathering initial evidence of a crisis and the recall. Both factors have already been empirically analyzed before. The findings, hitherto, are based on US data and do not seem to fit to our analysis of variance, which is based on a representative German sample of consumers. After this international comparison there will be a brief outlook containing remarks on managerial implications.