Martens [Soil Biol. Biochem. 33 (2001) 701] recently reported that Jenkinson and Oades [Soil Biol. Biochem. II (1979) 193] method for measuring adenosine 5? triphosphate (ATP) in soil leads to serious underestimations. We have now compared the Jenkinson and Oades extraction technique, which corrects for incomplete extraction of ATP by reference to a 'spike' of added ATP, with a sequential extraction procedure as used by Martens. Measurements of the ATP contents of four soils by the two procedures showed no significant differences, in contrast to Martens' findings: © 2002 Elsevier Science Ltd. All rights reserved.