AbstractAzolla pinnata R. Brown and its N‐fixing endophyte Anabaena azollae Strasburger are cultivated as a fallow‐season green manure for rice in Vietnam. This study presents field and laboratory experiments to determine if the temperate‐climate Azolla filiculoides Lam. could be grown in fallow flooded rice fields of California's Sacramento Valley.Field experiments in an agricultural drainage ditch indicate the exponential growth rate and N content of A. filiculoides is correlated logarithmically with maximum air temperature between early February and the end of April. Laboratory studies in controlled environment further revealed the exponential growth rate of A. filiculoides increases linearly with temperature between 10/1 and 25/15 C (12 hour thermoperiod and photoperiod) and remains high up to 35/25 C. Light saturation for growth is 100 µE✕m−2✕sec−1 at 10/1 C but at 35/25 C increases to at least 1,000 µE✕m−2✕sec−1. Nitrogenase activity increases with temperature between 10/1 C and 30/20 C, but is insensitive to light over a range of 50 to 1,000 µE✕m−2✕sec−1. Optimum N fixation for A. filiculoides is between 250 and 300 mg N✕g initial dry weigh−1✕week− and occurs at 25/15 to 30/20 C and 500 µE✕m−2✕sec−1.At 40/30 C the growth and nitrogenase activity of A. filiculoides is nil. However, if ferns are grown at lower temperatures and then subjected to a stepwise increase in temperature simulating dawn to midday of a diurnal cycle, nitrogenase activity increases with temperature up to 40 C and remains high at 45 C. Similar increases in nitrogenase activity with temperature are observed for field‐grown A. filiculoides during hot (40 to 45 C) afternoon periods.Inoculation of A. filiculoides equivalent to 1.2 kg N✕ha−1 onto 22.4 m−2 plots of a fallow flooded rice paddy results in N yields between 33 and 93 kg✕ha−1. Rice paddies used for 1977 studies were a Capay silty clay (Typic Chromoxererts), while 1978 studies were on a Hillgate loam (Typic Palexeralfs). On both soil series N yield of Azolla was correlated positively with the length of time the fern cover had developed prior to sporulation. Sporulation, which was followed by termination of nitrogenase activity and frond senescence, was hastened by high temperature. Thus, a mat of A. filiculoides developing in hot summer weather fixed only one‐third as much N as a population which developed over a longer period during early spring.Incorporation of 40 kg N✕ha−1 into soil as dry A. filiculoides in spring 1977 increased rice yield about 2.0 metric tons✕ha−1 over unfertilized controls. This effect was the same as would derive from an equal amount of N fertilizer in the form of ammonium sulfate and remained true when Azolla was supplemented with 40 and 80 kg N✕ha−1 as ammonium sulfate. Incorporation of 93 kg N✕ha−1 as dry A. filiculoides grown over a 46 day period in spring 1978 increased rice yield 2.6 MT✕ha−1, or 70%, of the increase obtained with an equivalent amount of ammonium sulfate.While more data are needed for A. filiculoides at different sites and years, results from this 2‐year study suggest up to 50% of the N requirement for rice in California could be supplied by one fallow‐season crop of Azolla. Further experiments will be needed to determine the optimum levels of inorganic nutrients (P and Fe) and labor inputs needed to grow Azolla before this source of biologically fixed N can be compared to N fertilizers currently used in temperate‐zone rice culture