The discovery of radiocarbon (14C) peaks in AD 774–775 and AD 993–994 sparked the search for other anomalous events, leading to the identification of one around 660 BC. However, the ∼660 BC event appears to show a more prolonged increase, raising the question whether the event is qualitatively different. To investigate this, we measured high-latitude tree rings from Finnish Lapland, expected to be highly sensitive to energetic particle events. We measured the 14C content of full rings, as well as their separated earlywood and latewood components. We found that the 14C concentrations start rising already in the latewood of 665 BC and reach almost its full intensity by 664 BC. This rapid increase is similar to that at another high-latitude site (Yamal, Russia) but contrasts with that of low-latitude sites, which show a later peak. The earlier increase of the 14C at high-latitude tree rings compared to lower latitudes is consistent with similar observations for the AD 774 and AD 993 Miyake events. Based on carbon-cycle box modeling, the structure of the subsequent amplitude increase can be explained by either single or double initial 14C pulses. The fast increase coupled with a slower subsequent peak structure suggests similar mechanisms behind the high-latitude observations, i.e., tropospheric 14C production and/or a fast component of polar air flow across the tropopause combined with the full stratospheric-tropospheric CO2 exchange. Our results strongly emphasize the need for dynamic carbon cycle models to understand the observed differences between high- and lower-latitude data.