A read-across assessment for oral systemic toxicity was undertaken for a series of nickel (Ni) substances. Bioaccessibility can provide information regarding potential bioavailability and subsequent toxicity such as acute and chronic oral toxicity and reproductive toxicity. Twelve Ni-containing samples were extracted in synthetic gastric and intestinal fluids. Acute oral toxicity studies in rats were conducted with eleven of these substances to verify that the bioaccessibility data provide reasonable estimates of systemic bioavailability and hence toxicity. The oral LD50 data were well predicted by Ni release data for each substance (R2=0.91). For substances with existing oral LD50 data, results were compared with their respective EU hazard classification. For substances without LD50 data, one sided lower confidence limits for predicted LD50 values were calculated. Results indicated with 95% confidence that samples releasing <48% available Ni (mg Ni released/mg available Ni x 100) are predicted to have an LD50 >2000 mg substance/kg; samples releasing >76% available Ni are expected to have an LD50 between 200-300 mg substance/kg; and samples with 48-76% bioaccessibility are predicted to have an LD50 between 300-2000 mg substance/kg. When Ni substances were allocated into these three groups according to bioaccessibility, the hazard classifications for systemic oral toxicity were determined based on read-across from three data-rich nickel compounds. Samples releasing <48% available nickel were read-across from Ni oxide and Ni subsulfide. Samples releasing >76% were read-across from Ni sulfate. No source or target substances released 48-76% available Ni in this study. Therefore, in the absence of additional verification data, read-across from the more conservative source substance (i.e., Ni sulfate) was recommended. Substances read-across from Ni sulfate warrant classification for acute (Xn;R22 and Acute Tox. 3) and reproductive (Cat 2;R61 and Repro 1B) toxicity in the EU.