A reexamination of original sources with references to earthquakes in New England and vicinity during the seventeenth century has revealed a number of discrepancies in presently available earthquake catalogs. Misinterpretations of earlier published reports by later catalog compilers as well as transcription errors have led to a number of false earthquakes being erroneously included in modern earthquake catalogs. Furthermore, three earthquakes that do not show up in most modern earthquake catalogs for the region have been found in some of the original diary sources. Earthquakes above magnitude 5 took place during the seventeenth century in northeastern North America in 1638, 1661, 1663 (5 events), 1665, and 1668 based on the widespread distances over which they were felt and on the strength and duration of the ground shaking. From the intensities where it was felt and a report of some aftershocks having been sensed in eastern Massachusetts, the 1638 earthquake is assigned an epicenter in central New Hampshire in this study with a magnitude of 6.5—7.0. The only other earthquakes for which epicenters are assigned in this study are the 1663 shocks in the Charlevoix region of Quebec and small earthquakes in 1668 (Littleton, Massachusetts), 1685 (Newburyport, Massachusetts) and 1697 (Brockton, Massachusetts). There is a description in one source of possible sandblows in Maine in 1670, but a contemporary source indicates that no earthquake accompanied the formation of those features. The features may have been caused by a small meteorite impact.