STUDY OBJECTIVE--The aim was to examine the relationship between handedness and longevity. DESIGN--This was an archival (retrospective) survey of a cohort of adult men who had played 'first-class cricket'. SETTING--The United Kingdom PARTICIPANTS--The subjects consisted of all of the deceased players included in an encyclopedia of 'first-class cricket' whose bowling hand had been recorded (n = 3165). The study also considered a further 2314 players, born before 1951 but still alive at the time the book was published (1984). MEASUREMENTS AND MAIN RESULTS--Using the bowling hand as an indicator of handedness it was possible to compare the lifespans of 2580 right handed men and 585 left handed men. The average life spans of the two groups differed by 25 months (right = 65.62, left = 63.52), a highly significant difference (p = 0.006). An examination of cause of death (where noted) strongly indicated that the left handed men were more likely to die prematurely in accidents or in warfare. As a consequence, when these unnatural deaths were removed from the sample the longevity difference between the right handers and left handers was considerably reduced. There was no evidence that these results related to any longitudinal change in the proportion of right handers to left handers across the time course of the sample. CONCLUSION--The study found clear evidence that left handedness was associated with a decrease in longevity among a cohort of adult, athletic men. A major factor responsible for this result seemed to be a differential likelihood of accidental death or death during warfare.