Complete dissolution is essential to obtain accurate analytical results for geological samples. Felsic rocks are known to be very difficult to dissolve because of the presence of refractory minerals such as zircon. In this study, we undertook a systematic evaluation of the effect of the HF/HNO3 ratio, digestion time, digestion temperature, digested test portion mass and the presence of insoluble fluorides on analytical results for the felsic rock GSP‐2 using high‐pressure HF and HF/HNO3 digestion. Digestion in mixtures of HF and HNO3 acids is a commonly used method of dissolution for geological samples. However, our results clearly indicate that adding HNO3 inhibited the digestion capabilities of HF for refractory minerals such as zircon. It took 8–12 hr for Zr to be completely recovered in GSP‐2 at 190 °C, whereas it needed about 36 and 72 hr at 160 and 140 °C, respectively. White precipitates were observed in the final solution for test portion mass > 100 mg, irrespective of which of the five different digestion solutions was used (1 ml HF, 2 ml HF, 1 ml HF + 0.5 ml HNO3, 1 ml HF + 1 ml HNO3 and 1.5 ml HF + 1.5 ml HNO3). Environmental scanning electron microscopy showed that these precipitates were mainly composed of AlF3. Instead of further HCl, HNO3 or HClO4 attack, we propose that using ultra‐fine samples and a small sample size is a good way to avoid the formation of insoluble residues (e.g., fluorides). To further investigate the precision and accuracy of the proposed method (using HF alone as the digestion solution during the first acid attack step), a suite of silicate rock reference materials was analysed. Most of the results were found to be in reasonable agreement with the reference values, with a relative error of < 10%.