A group of 576 leaf samples of cane, variety CB.41/76. was collected form a 2³ NPK factorial experiment in order to determine the most raliable sampling technique as a basis for the local conditions of the State of São Paulo. The selection of tissues for analysis was carried out on the plantation with respect to the age at sampling time (4, 5. 6, 7, 8 and 9 months in age), leaf part (blades, blades with midribs removed, and sheaths), and leaf position (leaves +1, +2, +3 and +4). This included the four youngest visible-dewlap leaves on the cane top, numbered according to the system of Kuijper. After each tissue was analyzed for N, P, K, Ca and Mg, the data were examined statistically, taking into account the response to the element supplied as fertilizer, as indicated by cane yields at harvest. As a result of these studies and considering the amount of work required in the laboratory tests and in taking samples in the field, it is suggested, for foliar analysis purposes, to select the leaf of position +3 in plant 4 and 8-9 months old: from the selected leaves the middle 8 inches of the leaf-blades with the midribs stripped furnish the most suitable material for total ntrogen, phosphorus, potassium, calcium and magnisium determinations. The use of one-sample technique at 4 months affords low operating costs. However, the use of two-sample technique is probably more suitable to indicate the nutritional status of sugar cane under the condition of soil and climate in the State of São Paulo, where the cane is usually harvested at 16-18 months.