Short wave near infrared (interactance geometry) spectroscopy is used in assessment of various attributes of intact fruit. As sample temperature impacts the water peak, the performance of fruit Brix (soluble sugar) and dry matter models is perturbed. Addition of a relatively small number of spectra of the same set of samples at different temperatures can create a model robust to temperature. However continued addition of samples at a uniform temperature can overwhelm this compensation effect, resulting in a model that is not robust to temperature. The aim of the current exercise is to document this effect, in order to propose population structures to ensure that a model remains robust to temperature. Use of orthogonal scatter correction (method of removal of a small number of factors which account for large variation in spectra and are orthogonal to the reference value; Fearn, 2000) and of a repeatability file (Westerhaus, 1990) was also considered. Models were developed using spectra of tomato fruit collected at 15°C, and tested on spectra of an independent set of fruit at a sample temperature of 35°C.