The Thellier technique is the most commonly used method for absolute palaeo
intensity studies. It is based on Thellier's laws stating that partial ther
moremanent magnetisations (pTRMs) are independent and additive, which howev
er only are true for non-interacting single domain (SD) grains. For multido
main (MD) grains, individual pTRMs are not independent and Thellier palaeoi
ntensity experiments on rocks with a significant proportion of MD grains ma
y therefore lead to erroneous palaeofield estimates. By introducing a "pTRM
-tail check" in the Coe version of the Thellier technique [J. Geophys. Res.
72 (1967) 3247], we present a simple method to detect and discard such roc
ks. The usefulness of the pTRM-tail check is exemplified by Thellier palaeo
intensity experiments performed on two palaeomagnetic sample collections wi
th different magnetomineralogy. Miocene baked sediments from French Massif
Central are shown to be ideal for Thellier studies with the remanent magnet
isation residing in thermally stable SD grains. For Palaeocene-Eocene basal
tic lavas from Faeroe Islands, 64% of otherwise acceptable palaeointensity
estimates are discarded due to the failure of pTRM-tail checks indicating t
he presence of significant concentrations of MD grains. (C) 2001 Elsevier S
cience B.V. All rights reserved.