Fine root (<= 1.0 mm diameter) respiration and longevity were studied in four sugar maple (Acer saccharum Marsh.) dominated northern hardwood forests. The sites were located along a latitudinal temperature gradient and also differed in N availability. Root respiration was measured on excised roots as O2 consumption at temperatures from 6 to 24 C, and root longevity was assessed by direct observation using minirhizotrons. The experiments were conducted across multiple years which differed significantly in moisture availability. Fine root respiration at the four sites was exponentially related to soil temperature (Q10=2.7) and linearly related to root N concentration and soil moisture availability. Temperature explained 90% of the variability in respiration. Differences among sites in respiration rates resulted from site-specific differences in N availability and root N concentration, and differences in soil moisture availability explained temporal variation within sites in respiration rate at a given temperature. Periodic moisture deficits were sufficient to cause declines of up to 17% in total growing season root respiration. Root respiration increased significantly as the [CO2] at which measurements were made was lowered, and was most sensitive to [CO2] near and below normal soil concentrations (<1500 ul/l). This suggests that estimates of root respiration made at or near atmospheric [CO2] may overestimate actual rates occurring in the field. Sugar maple root respiration at atmospheric [CO2] (350 ul/l) was about 139% of that at soil [CO2] (1500 ul/l). There were significant differences in surface root longevity among the sites, with median root lifespans for spring 1994 roots ranging from 278 to 397 days. The pattern across sites in root longevity corresponded to site differences in N availability, with longer root lifespans occurring where N availability was greater. It appears that roots are maintained as long as the benefit (nutrients) they provide outweighs the C cost of keeping them alive. It is hypothesized that greater root metabolic activity in nitrogen rich zones leads to greater carbohydrate allocation to the roots, and that reduction in root C sink strength, as local nutrients are depleted, provides the mechanism through which root lifespan is regulated.