The temperature-programmed desorption (TPD) of N-2 from a Ru/MgO catalyst used for ammonia synthesis was studied in a microreactor flow system operating at atmospheric pressure. Saturation with chemisorbed atomic nitrogen (N-*) was achieved by exposure to N-2 at 573 K for 14 h and subsequent cooling in N-2 to room temperature. With a heating rate of 5 K/min in He, a narrow and fairly symmetric Na TPD peak at about 640 K results. From experiments with varying heating rates a preexponential factor A(des) = 1.5 X 10(10) molecules/(site s) and an activation energy E(des) = 158 kJ/mol was derived assuming second-order desorption. This rate constant of desorption is in good agreement with results obtained with a Ru(0001) single crystal surface in ultra-high vacuum (UHV). The rate of dissociative chemisorption was determined by varying the N-2 exposure conditions. Determination of the coverage of N-* was based on the integration of the subsequently recorded Ng TPD traces yielding A(ads) = 2 x 10(-6) (Pa s)(-1) and E(ads) = 27 kJ/mol. The corresponding sticking coefficient of about 10(-14) at 300 K is in agreement with the inertness of Ru(0001) in UHV towards dissociative chemisorption of N-2. However, if the whole catalytic surface were in this state, then the resulting rate of N-2 dissociation would be several orders of magnitude lower than the observed rate of NH3 formation. Hence only a small fraction of the total Ru metal surface area of Ru/MgO seems to be highly active dominating the rate of ammonia formation.