Abstract A detailed palaeomagnetic investigation of early Silurian volcanics of the Lawrenceton Formation in the Central Mobile Belt (CMB) of Newfoundland reveals primary remanence that places the sampled part of the CMB on the margin of Laurentia in low southerly latitudes in the early Silurian. The palaeomagnetic data indicate large vertical axis (net clockwise) rotations within the CMB and between the CMB and cratonic Laurentia. The volcanics were folded, faulted and in places cleaved in Silurian to Early Devonian times. Anisotropy of magnetic susceptibility measurements on lava flow samples verify field observations of a primary flow-parallel foliation and confirm field measurements of the palaeo-horizontal. A well-developed tectonic fabric was detected at one sampling site near the Reach Fault. Silurian redbeds of the Big Indian Pond Formation were also sampled for palaeomagnetic study. The sediments are folded, faulted and cleaved. A statistically sound syn-folding remanence was identified which, in a tectonically corrected state, is consistent in direction with the remanence in the Silurian volcanics. The sampled sediments are more intensely deformed than the volcanics and magnetic anisotropy measurements indicate both primary layering- and secondary cleavage-related fabrics in varying degrees from site to site. It is possible that cleavage development resulted in remanence reorientation leading to the illusion of syn-folding remanence when, in fact, the remanence was acquired prior to folding. Previously published palaeomagnetic investigations of Silurian rocks in the CMB revealed a discrepancy in palaeofield inclination between volcanic and sedimentary rock types. The sedimentary rocks yielded shallow inclinations and the volcanics intermediate inclinations. Inclination shallowing in the sediments was put forward as an explanation, as well as contamination of the sedimentary results with a shallow overprint. Others argued that the results from the volcanics were in error. The present investigation demonstrates that key volcanics carry a shallower remanence than previously thought, bringing them into approximate agreement with the majority of the results from sediments. Steep results from the Springdale volcanics and very shallow results from the King George IV lake sediments are at variance with the rest of the data. The Springdale result has large error limits that overlap with the other data. The anomalous King George IV result remains an outlier.