Ichthyoneuston surveys were conducted using a manta neuston sampler at five fixed stations along a transect, 1-19 km from shore, off Davenport, CA from December 1991-June 1992, October 1992, and January-April 1993. Larval and juvenile fishes were collected. Five taxa, Engraulis mordax, Atherinopsis californiensis, Sebastes spp., Merluccius productus, and Hexagrammos spp. (predominantly H. decagrammus) made up 94% of the mean density of total ichthyoneuston (859/1000 m3). Atherinopsis californiensis and Hexagrammos spp. were obligate ichthyoneuston. Mean density of total ichthyoneuston from January-April was two times greater in 1992 (1485/1000 m3) than in 1993 (689/1000 m3) and is attributed to the decreased abundance of 11 of 18 dominant taxa in 1993. Ichthyoneuston taxa formed nearshore, cosmopolitan, and offshore assemblages that generally remained cohesive between years as indicated by cluster analyses. However, distributions of abundant taxa within the cosmopolitan assemblage shifted offshore to varying extent in 1993. Separate cluster analyses of samples from January-April of each year indicated that species composition of samples from the first two or three months differed from samples of later months, and the shift from an "early" to "late" assemblage occurred earlier in 1993 (March) than in 1992 (April). Common obligate ichthyoneuston taxa, except for Scorpaenichthys marmoratus, were most abundant in samples from the first two or three months during both years. Regional wind patterns and local oceanography (e.g., prolonged wind reversal event in February 1992; earlier onset of persistent upwelling-favorable winds in 1993; substantial fresh water influence in 1993) may contribute to between-year differences in the temporal and spatial distribution of ichthyoneuston.