The seasonal change of the rates of production and emission of methane were determined under in-situ conditions in an Italian rice paddy in 1985 and 1986. The contribution to total emission of CH4 of plant-mediated transport, ebullition, and diffusion through the flooding water was quantified by cutting the plants and by trapping emerging gas bubbles with funnels. Both production and emission of CH4 increased during the season and reached a maximum in August. However, the numbers of methanogenic bacteria did not change. As the rice plants grew and the contribution of plant-mediated CH4 emission increased, the percentage of the produced CH4 which was reoxidized and thus, was not emitted, also increased. At its maximum, about 300 ml CH4 were produced per m2 per hour. However, only about 6% were emitted and this was by about 96% via plant-mediated transport. Radiotracer experiments showed that CH, was produced from H2/CO2. (30–50%) and from acetate. The pool concentration of acetate was in the range of 6–10 mM. The turnover time of acetate was 12–16 h. Part of the acetate pool appeared to be not available for production of CH4 or CO2