Carbon dioxide evolution rates for downed logs (coarse woody debris) and the forest floor were measured in a temperate, old-growth rain forest in Olympic National Park, Washington, using the soda lime trap method. Measurements were taken every 4 weeks from October 22, 1991, to November 19, 1992. Respiration rates for Douglas-fir (Pseudotsugamenziesii (Mirb.) Franco) and western hemlock (Tsugaheterophylla (Raf.) Sarg.), logs were determined for decay classes 1–2, 3, and 5 in two diameter classes. Overall, western hemlock logs respired at a rate 35% higher (4.37 g CO2•m−2•day−1) than Douglas-fir logs (3.23 g CO2•m−2•day−1). Respiration rates for decay class 1–2 logs of both species were similar to decay class 5 logs (4.46 and 4.07 g CO2•m−2•day−1, respectively), but decay class 3 logs respired at a lower rate (3.23 g CO2•m−2•day−1). Seasonal patterns of respiration rates occurred, particularly for decay class 1 and 2 western hemlock logs where monthly averages ranged from a low of 2.67 g CO2•m−2•day−1 in February 1992 to a high of 8.30 g CO2•m−2•day−1 in September 1992. Rates for decay class 1–2 western hemlock logs were greater than those from the forest floor, which ranged from 3.42 to 7.13 g CO2•m−2•day−1. Respiration rates were depressed in late July and August compared with fall and spring owing to the summer drought characteristic of the Pacific Northwest. Large-diameter western hemlock logs in decay class 1–2 had higher respiration rates than small-diameter logs, whereas large-diameter decay class 3 western hemlock logs had lower respiration rates than small-diameter logs.