This paper presents the results of an inter-comparative study in view of assessing the reliability of radiocarbon dates obtained on calcined bones from open-air Palaeolithic and Mesolithic sites. The results demonstrate that the success rate is largely dependent on site-taphonomy, in particular the speed of covering of the site. Sites quickly covered by aeolian, alluvial or marine sediments yield on average good dating results. At worst they can be affected by an wood-age offset, generally <100 years, caused by the uptake of carbon from the firewood. Sites which are uncovered or have been covered rather late suffer from contamination problems resulting in radiocarbon dates much younger than the reference dates. For these sites, which unfortunately represent the vast majority of open-air Palaeolithic and Mesolithic sites, calcined bones are not a valuable dating material for developing robust, decadal-to-centennial chronologies.