RATIONALE: We investigated the applicability of tree-ring whole-wood material for δ(18)O and δ(13)C analysis in comparison with the more time- and resource-intensive use of cellulose, by considering possible variability between (i) five different tree species (Fagus sylvatica, Quercus robur, Picea abies, Abies alba, Pseudotsuga menziesii), (ii) two sites that differ in soil moisture, and (iii) climate conditions within a 10-year period. METHODS: Stem cores of 30 individual trees (n = 3 trees per each species and site) were sampled from two sites in south Germany (Bavaria), and tree rings within sapwood of the years 2001-2010 were separated. The δ(18)O and δ(13)C values from homogenized tree-ring whole wood and from extracted cellulose were measured by mass spectrometry. Species-specific offsets in isotope values were analyzed and the responses in isotopic signature to climate variability including a single drought event were compared between whole-wood and cellulose. RESULTS: A constant offset in δ(18)O values of ca 5‰ between wood and cellulose was observed for most species independent of site conditions, with a significant difference between beech and Douglas-fir, while inter-annual variability was only observed in oak. The offset in δ(13)C values ranged between 1.45 and 1.84‰ across species, sites and years. Both materials generally showed similar strength in responses to temperature, precipitation and soil water availability, particularly for conifers. Resistance to severe drought stress--partly more strongly reflected in the δ(13)C values of cellulose--was lower for conifers than for the deciduous species. CONCLUSIONS: Wood material from the sapwood of the studied tree species is as useful as cellulose for studying environmental effects on tree-ring δ(18)O and δ(13)C values at a short-term scale as considered in most ecophysiological studies. The more variable response of oak may require further investigations.