Citrus is a major crops in Brazil, especially in the São Paulo state, which is the main citrus production region in the world. Yet, site specific technology is still in early stages of adoption. Variable rate application of inputs is the most important tool in a Precision Agriculture system, however its effect on citrus agronomical aspects are still unknown, especially during long periods of observation. Thus, variable rate fertilizer application has been tested in citrus orchards in Brazil in a long term study. The goal is to evaluate the effects of this technology on input consumption, soil fertility, plant nutrition and on fruit yield. Two 25.7 ha commercial orange fields (field 1, 32% clay and field 2, 14% clay) were divided into variable and fixed rate fertilizer strip treatments. Data of soil electrical conductivity, elevation and soil texture were used to assess the variability of these fields. Variable rate prescriptions (lime, N, P and K) were based on soil and leaf grid sampling and yield data. Fixed rate applications followed standard prescriptions based on soil sampling and yield expectation. Results from the first two years (2008 and 2009) were presented at the last ICPA, in 2012 (Colaço and Molin, 2012). Data from three more years were gathered and analyzed, accomplishing so far five out of six year data planned for this study. Results presented here are from five yield data (2008 until 2012) and four variable rate fertilizations (2008 until 2011). Field 1 showed more variability than field 2 regarding soil texture, EC and elevation. Site specific applications provided significant reduction on input consumption, mainly for nitrogen (37 and 51% less on field 1 and 2, respectively) and potassium fertilizers (41 and 18% less on field 1 and 2, respectively). In field 1, better fertility levels were found on the variable rate treatment. Along the years of evaluation, it reduced regions with excess of nutrients and enlarged areas of adequate levels of potassium and base saturation. Yield gains up to 13.1% occurred on this field. In the second field, loss on soil fertility and yield was found on the site-specific management in two out of the four years of variable rate applications. This field presented lower natural nutrient fertility and the variable rate prescriptions used were considered not suited for this soil condition. Fertilizer use efficiency (fruit yield divided by fertilizer input) was higher in the variable rate treatment in both fields. Leaf nutrition was not affected by the treatments. Overall results were better for the field with higher variability. Site specific management allowed optimization and higher fertilizer use efficiency. This study showed the potential of this technology to increase yield and improve soil fertility management.