Four California annual grassland species were used to examine the hypothesis that different plant species have equivalent competitive effects. We investigated the effects of the annuals on soil water availability and the growth responses of blue oak (Quercus douglasii Hook&Arn.) to neighbor—induced water depletion. Neighborhoods of annuals were composed of species from California annual grassland with differing phenology and morphology that were hypothesized to show non—equivalent competitive effects on both a per—individual— and a per—unit—tissue basis. Three introduced species of winter annuals were sown at each of three densities (10, 30, or 100 seeds/dm2) around a single oak acorn. The grass Bromus diandrus Roth. and the forb Erodium botrys Cav. flower in early spring, while the grass B. mollis L. flowers slightly later. A native summer annual forb, Hemizonia luzulaefolia DC., was sown at only the intermediate density. Plants were grown outdoors in 15 cm diameter X 1 m deep tubes filled with soil from an oak woodland site. Identity and density of annual species had independent and interactive effects on the water resource level. Soil water potential was negatively density—dependent, remaining highest when oaks were grown without competitors. At the intermediate density soil water potentials decreased in the following order: B. mollis (slowest), B. diandrus, E. botrys, and H. luzulaefolia (most rapid). This order followed the pattern of root length developed in the neighborhoods, which was species— and density—dependent. The relationship between sowing density and neighborhood dry mass also differed among the annual species. Correlations between root length and soil water potential were stronger for the forbs than for the grasses. A given root length of the summer annual, H. Luzulaefolia, depleted soil water to a greater extent than did the same root length of any species. Oak seedling growth, gas exchange rates, and survivorship were negatively related to the rate and extent of soil water depletion. Additionally, the relationship between annual root length and oak seedling height or root length differed by neighborhood species. Thus, neighborhood species were non—equivalent in competitive effect on both a per—individual and a per—unit—root—length basis. Phenologically different species may represent different functional groups of competitors.