The utilization of Eucalyptus spp. in Uruguayan pulp mills for their production isbalready well established, and therefore the characterization of the wood is one of the first steps to evaluate the process conditions or the suitability of the raw material. However, most of the conventional methods are time-demanding and usually have high costs. The Near Infrared Spectroscopy (NIRS) methodology has been developed to avoid these problems, rapidly, with a precision according to the needs and comparatively low cost. There are a significant number of properties that can bebanalyzed in a pulp, and we have selected three for this work. The acetone extractives determination involves measuring substances as fatty acids, resin acids, sterols, waxes and non-volatile hydrocarbons. The selection of woods for cellulose pulp production is done considering those species having low acetone-extractable content. Also the quantity found in a sample depends on seasoning or drying of wood (Tappi T 280, 1999). The presence of wood extractives can raise the chemical reagent consumption and produce foam during pulp wash (Núñez, 2008). The one percent sodium hydroxide solubility of wood generally indicates the degree of a fungus decay or degradation by heat, light or oxidation: an increase in the solubility means a higher decay or degradation in wood. The principal components remaining in the hot alkali solution are hemicelluloses and degraded cellulose, and therefore the determination is also an indicator of the presence of lignin (Tappi T 222, 2006). The third determination that was calibrated by the NIRS technique was the acid insoluble lignin or Klason lignin, which represents 80 % to 90 % of the total lignin (Hausalo and Söderhjelm, 2000; Sixta 2006). In the NIR technique, the first step consists of a determination by the conventional method of a great number of representative samples. Next, it involves the measurement of the spectra of those samples and the development of a model that can quantify the variations in the properties. Finally, this model is used to predict the value of an unknown sample by measuring the near infrared spectra. This work, a continuation of the previous calibration developed for wood basic density and pulp yield (Bonfiglio et al., 2011), shows the calibration results for those three chemical properties previously discussed that are usually demanded.