In this study, an experimental design was conceived, as part of the semi-Arid-Land-Surface-Atmosphere (SALSA) program, to document the effect of view angle variation on surface radiative temperature measurements. The results indicated differences between nadir and off-nadir radiative temperature of up to 5K. The data also illustrated that, under clear sky and constant vegetation conditions, this difference is well correlated with surface soil moisture. However, the correlation decreased when the same comparison was made under changing vegetation conditions. To investigate the possibility of deriving component surface temperatures (soil and vegetation) using dual-angle observations of directional radiative temperature, two radiative transfer models (RTM) with different degrees of complexity were used, The results showed that despite their differences, the two models performed similarly in predicting the directional radiative temperature at a third angle. In contrast to other investigations, our study indicated that the impact of ignoring the cavity effect term is not very significant. However, omitting the contribution of the incoming long-wave radiation on measured directional radiance seemed to have a much larger impact. Finally, sensitivity analysis showed that an accuracy of better than 10% on the plant area index (PAI) was required for achieving a precision of 1K for inverted vegetation temperature. An error of 1K in measured directional radiative temperature can lead to an error of about 1K in the soil and vegetation temperatures derived by inverting the RTM.