Experiments in agriculture usually consider topsoil properties uniform and, for this reason, often make bad use of the results. The soil water content has spatial variability due to other proprieties. Then it is necessary to know how this variation occurs in the spatial and time. The objective of this study was to assess the spatial variability as well as the temporal stability of spatial distribution for soil moisture content. The experiment was carried at Central Experimental Farm, of Instituto Agronômico (IAC), SP, Brazil, in an area of 3.42 ha, from August 2003 to January 2004. The soil was classified as Rhodic Ferralsol (typic Haplorthox). This area was divided in a 10 m grid with 302 points, from which a rectangular grid with 10m spacing for direction X and 20m for Y was laid out, resulting in 102 sampling points. Access tubes were inserted 1m deep at each sampling point to receive a neutron moisture gauge for measurements at 30, 60 and 90 cm depth. Measure-ments were taken on 7 occasions between the end of the winter and beginning of the spring of 2003 and 4 occasions in January 2004. The data for soil moisture content (cm3cm-3) were analyzed using geostatistical techniques. The spatial vari-ability was analyzed examining the scaled semivariograms, the statistical parame-ters (mean, variance, coefficient variation, minimum value, maximum value, skewness, kurtosis) and the parameters of the models fit to the individual semivariograms as a function of time. In general, semivariograms show that all data sampled have strong spatial dependence to about 90 and 110m and this de-pendence exists only at the 60 cm depth. It was concluded that the spatial variability of the soil moisture content increased from winter to summer, when the soil has a higher water content, with a smaller coefficient of variation of the data.