A method for the determination of titanium in graphite furnace atomic absorption spectrometry with slurry sampling was developed. Titanium forms thermally stable carbides in the graphite tube that leads to decreased sensitivity and severe memory effects. Various fluorinating agents, BaF(2), NH(4)F, and CHF(3) (Freon-23) were therefore examined in order to reduce or eliminate these problems. Ti was determined, at various concentration levels, in certified reference materials (CRMs) using ultrasonic slurry sampling graphite furnace atomic absorption spectrometry (USS-GFAAS). The three CRMs, GBW 07601 (Human Hair Powder), GBW 07602 (Bush Branches and Leaves), and GBW 07411 (Chinese Soil), contained 2.7 microg g(-1), 95 microg g(-1), and 0.41% Ti, respectively. For comparison, determinations of Ti were made with modifiers (BaF(2) and NH(4)F) and without modifier, using 5% CHF(3) (in argon) for cleaning the graphite furnace. Good accuracy was obtained using aqueous Ti standards for calibration. A homogeneity study showed that Ti was evenly distributed in all the samples at the mg-microg level. The relative standard deviations (RSDs) obtained for the three CRMs were 16%, 11%, and 8% ( n=30). In spite of the wide range of Ti concentrations in the present samples, the same wavelength (365.4 nm) could be used for analysis by varying the slurry sample concentration. The precision was best for the material with the highest titanium content in spite of the fact that only 3 microg of sample was introduced into the furnace.