Knowledge regarding the historic role of fire in forest ecosystems is often derived by dating fire scars. Fire history studies in boreal and western coniferous forests have used either complete or partial cross sections cut from fire-scarred trees. The occurrence of missing or false rings presents an obstacle to the accurate dating of individual fires. Two different techniques of dating partial cross sections of fire-scarred ponderosa pine (Pinus ponderosa Laws) were compared using the same samples. Method I involves the tabulation of raw dates followed by a subsequent adjustment based on synchronization with fire dates from adjacent trees. Method II uses correlations with master chronologies based on analysis of increment cores taken in the same region as the sampled fire scar sections. Method I is prone to error since the assumption that fire dates differing by a few years may actually be from the same year cannot be independently verified. Method II provides a higher degree of certainty with regard to the accuracy of individual dates because the cross dating of local patterns of ring widths circumvents the problem of absent or false rings. Eight samples from the Horse Pasture Plateau in Zion National Park, Utah, were dated by both methods. Agreement on dates derived by the two methods for 39 separate fire years was limited to only 26 percent. The average error of Method I was 1.0 year. Fire history studies using Method I are accurate enough for most managerial recommendations and some ecological interpretations, but they may lack the precision necessary to correctly identify fire years or differentiate small changes in fire frequency. The dendrochronological approach of Method II is preferable because of its greater reliability, especially in ecosystems where short fire intervals are common or where fires occur in consecutive years. FORESX SCI. 28:856861. ADDITIONAL