Ti separation was achieved by ion-exchange chromatography using Bio-Rad AG 1-X8 anion-exchange and DGA resins. For high-Fe/Ti and high-Mg/Ti igneous samples, a three-column procedure was required, whereas a two-column procedure was used for low-Fe/Ti and low-Mg/Ti igneous samples. The Ti isotopes were analysed by MC-ICP-MS, and instrumental mass bias was corrected using a 47Ti-49Ti double-spike technique. The 47Ti-49Ti double-spike and SRM 3162a were calibrated using SRM 979-Cr, certificated value 53Cr/52Crtrue = 0.11339. Isobaric interference was evaluated by analysing Alfa-Ti doped with Na, Mg, Ca, and Mo, and results indicate that high concentrations of Na and Mg have no significant effect on Ti isotope analyses; however, Ca and Mo interferences lead to erroneous δ49/47Ti values when Ca/Ti and Mo/Ti ratios exceed 0.01 and 0.1, respectively. Titanium isotopic compositions were determined for 12 igneous reference materials, BCR-2, BHVO-2, GBW07105, AGV-1, AGV-2, W-2, GBW07123, GBW07126, GBW07127, GBW07101, JP-1, and DTS-2b. Samples yield δ49/47Ti (‰) of ?0.035 ± 0.022, ?0.038 ± 0.031, 0.031 ± 0.022, 0.059 ± 0.038, 0.044 ± 0.037, 0.000 ± 0.015, 0.154 ± 0.044, ?0.044 ± 0.018, 0.010 ± 0.022, 0.064 ± 0.043, 0.169 ± 0.034, and ?0.047 ± 0.025 (relative to OL-Ti, ±2SD), respectively; of which isotopic compositions of DTS-2b, JP-1, GBW07101, GBW07105, GBW07123, GBW07126, and GBW07127 are reported for the first time. Standard Alfa-Ti was analysed repeatedly over a ten-month period, indicating a reproducibility of ±0.047 (2SD) for δ49/47Ti, similar to the precisions obtained for geochemical reference materials.