In the first part of this study we analysed the temporal stability of Garfield’s Impact Factor (IF) over the last decade (2001-2010). The analyses were performed for two Web of Science (WoS) categories of the JCR Science Edition (“Polymer Science” and “Nanoscience & Nanotechnology”) and two of the Social Sciences Edition (“Political Science” and “Information Science & Library Science”) at journal category level. Furthermore, we tried to identify the most characteristic patterns of IF timelines at journal level and analysed the reasons for strong IF fluctuations. Additionally, we checked if alternative journal impact measures like the Article Influence Score, SJR or SNIP are as sensitive to short-term fluctuations in the citation frequencies as Garfield’s Impact Factor. In the second part, we explored if one often mentioned weakness of the IF, its short citation window, can be used to identify hot papers, i.e. papers that are cited clearly above-average in the first two years after publication, by means of IF fluctuations. By analysing the citation distributions at article level we show that abrupt and large short-term variations of the IF can, in principle, offer a very simple and intuitive method to identify “unexpected” hot papers and hot topics at journal level.