In this work, the possibilities of electrothermal vaporization-inductively coupled plasma mass spectrometry for the direct determination of B in biological samples have been critically evaluated, the goal always being to develop a fast and robust method that could be applied to different samples without significant modifications to the working protocol. Four reference materials of different nature (NIST SRM 1570a Trace Elements in Spinach Leaves, NIST SRM 1573a Tomato leaves, NIST SRM 1548a Typical Diet and BCR CRM 281 Rye Grass) were selected for this study. The importance of the use of NH₄F·HF as chemical modifier for this element was demonstrated. By using this modifier, it was observed that the analyte can be vaporized and transported to the ICP in a more efficient way even at low vaporization temperatures (from 1000 °C onwards), resulting in a sensitivity increase (of a factor of approximately 10 compared with the vaporization in the absence of any modifier) while minimizing memory effects and carbon co-vaporization.The temperature program was adapted to permit the efficient use of this modifier when direct analysis of solid samples was intended. However, although well-defined B signal profiles could be attained also for the solid samples, evidence of signal suppression was observed. Under these conditions, the use of external calibration with aqueous standards would result in underestimated results (40–60% of the expected values). It was demonstrated that the use of isotope dilution can satisfactorily correct for these suppression effects, allowing accurate results to be obtained for all the samples under investigation. The most relevant figures of merit of the method finally developed are: RSD values ranging from 7 to 9%, a detection limit of 8 ng g⁻¹, a sample throughput of 20–25 min per determination and a sample consumption of approximately 5 mg per determination.