Previous studies have shown controversial results on whether acetylator status causes isoniazid-induced hepatotoxicity (IIH). Moreover, the contribution of CYP2E1, a hepatic enzyme implicated in the formation of hepatotoxins, to the risk of developing IIH remains unclear. The objectives of this study were (i) to assess the quantitative relationship between the level of isoniazid serum concentration and the incidence of IIH and (ii) to evaluate the extent of implication of the N-acetyltransferase-2 (NAT2) and CYP2E1 polymorphisms genes to induce this disorder. Seventy-one patients with tuberculosis receiving a conventional antituberculosis regimen were included. NAT2 and CYP2E1 genotypes were determined using polymerase chain reaction. Three restriction enzymes, RsaI, PstI and DraI were used to detect CYP2E1 RFLP and four different restriction enzymes, KpnI, TaqI, BamHI and Ddel were used to determine NAT2 acetylator status. Therapeutic drug monitoring (TDM) of isoniazid (serum concentration performed 3 h after the morning dose: C3) was performed. Cases of isoniazid-induced hepatotoxicity were diagnosed according to Benichou et al. Receiver Operating Characteristics curve analysis was used to evaluate the relationship between risk factors and the incidence of IIH. Eleven (15.4%) patients have developed IIH. Demographic factors, including age, weight and gender were not associated with the incidence of hepatotoxicity. High serum concentration of isoniazid (C3) was found to be a risk factor of IIH (area under the curve: 0.74, P = 0.007, 95% confidence interval (95% CI): 0.56-0.93), with a cutoff value at 3.69 mg l-1 (odds ratio (OR): 13.2, P = 0.0007, 95% CI: 2.9-59). Multivariate analysis showed that only a C3 over 3.69 mg l-1 remains a risk factor of IIH. NAT2 and CYP2E1 variants were not found to increase the risk of IIH when analyzed separately. However, combined analysis of the NAT2/CYP2E1 gene polymorphisms showed that patients with both DraI C/D and slow acetylator have an increased risk of IIH compared with other combined NAT2/CYP2E1 genotype profiles (OR: 8.41, P = 0.01, 95% CI: 1.54-45.76). Our results suggest that a serum concentration of isoniazid over 3.69 mg l-1 and a combined genotype CYP2E1 DraI(C/D)/slow acetylator are major risk factors for IIH. Therefore, TDM of isoniazid and the determination of both NAT2 and CYP2E1 genotypes could be useful for the prediction and prevention of IIH in Tunisian tuberculosis patients.