The remanent magnetization of the Upper Devonian Perry Formation of New Brunswick consists of two components B and A acquired before and after folding (pre-Westphalian D) respectively. The stability spectra of these components overlap and it is therefore difficult to define them accurately and calculate their errors. The best estimate of the older (B) component (with respect to bedding) is 175,+23 (pole 32° N, 118° E), but this may be in error by 15° or more: its age is considered to be Late Devonian or Mississippian. The best estimate of the younger (A) component (with respect to present horizontal) is also 175, +23 and its age is considered to be early Pennsylvanian or older. The uncertainty in estimating these directions, however, means that the paleomagnetic observations, although they do not preclude the possibility that Newfoundland has rotated, also do not provide independent evidence that this has occurred, as has been suggested.