Geo-statistical modelling of rubber growing soils of Kottayam district in Kerala was carried out to assess soil fertility status of rubber plantations in order to develop a WebGIS enabled platform for soil fertility recommendation to rubber farmers. About two thousand and eighty two soil samples were collected from rubber growing areas of the district on a fifty hectare grid basis using satellite-derived rubber distribution maps. Soil samples were analysed for soil pH, organic carbon, primary, secondary and micronutrients following standard procedures. Geo-statistical modelling was performed to interpolate soil fertility parameters using ordinary kriging algorithm in GIS platform to generate soil fertility maps. Best fit semi-variogram models were used to interpolate soil fertility parameters and cross-validated with standard prediction error parameters. Nugget-to-sill ratios of the semi-variogram model revealed that degree of spatial autocorrelation of most of the soil fertility parameters were moderate in the study area. Spatial variability of soil fertility parameters indicated that major portion of NR growing soils in the study area exhibited acidic status of soil pH. Available organic carbon status was high to very high whereas status of available phosphorus and potassium were low. Secondary nutrients such as available calcium, magnesium and sulphur were also exhibited low status. Micronutrients status was sufficient but available boron and zinc showed deficient in some places of the district. A WebGIS enabled application called Rubber Soil Information System (RubSIS) was developed using the soil fertility variability maps and soil depth data of the study area. It gives location-specific and need-based recommendation for use of chemical fertilizers in one holding according to the age and the extent of the rubber plantation. Results of geo-statistical modelling of soil fertility parameters used to develop RubSIS are briefly discussed in the paper.