We have applied a non-contact method for studying the temperature changes produced by radiofrequency (RF) radiation specifically to small biological samples. A temperature-dependent fluorescent dye, Rhodamine B, as imaged by laser scanning confocal microscopy (LSCM) was used to do this. The results were calibrated against real-time temperature measurements from fiber optic probes, with a calibration factor of 3.4% intensity change degrees C(-1) and a reproducibility of +/-6%. This non-contact method provided two-dimensional and three-dimensional images of temperature change and distributions in biological samples, at a spatial resolution of a few micrometers and with an estimated absolute precision of around 1.5 degrees C, with a differential precision of 0.4 degree C. Temperature rise within tissue was found to be non-uniform. Estimates of specific absorption rate (SAR) from absorbed power measurements were greater than those estimated from rate of temperature rise, measured at 1 min intervals, probably because this interval is too long to permit accurate estimation of initial temperature rise following start of RF exposure. Future experiments will aim to explore this.