Pre-encounter ground-based thermal observations of NEA 25143 Itokawa at 10μm led to a size prediction of 520(±50) × 270(±30) × 230(±20)m, corresponding to an effective diameter of DTPMeff= 318m (Müller et al.2005). This is in almost perfect agreement with the final in-situ results 535 × 294 × 209m (DHayabusaeff= 320m; Demura et al.2006). The corresponding radar value, based on the same shape model (Kaasalainen et al.2005), were about 20% too high: 594 × 320 × 288m (DRadareff= 379m; Ostro et al.2005). The very simple N-band observations revealed a surface which is dominated by bare rocks rather than a thick regolith layer. This prediction was nicely confirmed by the Hayabusa mission (e.g., Fujiwara et al.2006; Saito et al.2006). The ground-based measurements covered three different phase angles which enabled us to determine the thermal properties with unprecedented accuracy and in excellent agreement with the results from the touch-down measurements (Okada et al.2006; Yano et al.2006). These thermal values are also key ingredients for high precision Yarkovsky and YORP calculations (mainly the rotation slowing) for Itokawa (e.g., Vokrouhlický et al.2004; Vokrouhlický et al.2005). In addition to the above mentioned properties, our data allowed us to derive the surface albedo and to estimate the total mass. We believe that with our well-tested and calibrated radiometric techniques (Lagerros 1996,1997,1998; Müller & Lagerros 1998, 2002; Müller 2002) we have tools at hand to distinguish between monolithic, regolith-covered and rubble pile near-Earth objects by only using remote thermal observations. This project also emphasizes the high and so far not yet fully exploited potential of thermophysical modeling techniques for the NEA/NEO exploration.