Background Poor response to TB therapy might be attributable to subtherapeutic levels in drug-compliant patients. Pharmacokinetic parameters can be affected by comorbidities or the interaction of drugs with food. Objectives This study aimed to determine the effect of food intake upon pharmacokinetics of rifampicin and isoniazid in a Peruvian population with TB. Methods Rifampicin and isoniazid levels were analysed at 2, 4 and 6 h after drug intake in both fasting and non-fasting states using LC-MS methods. Results Sixty patients participated in the study. The median rifampicin C max and AUC 0-6 were higher during fasting than non-fasting: 7.02 versus 6.59 mg/L (P = 0.054) and 28.64 versus 24.31 mg·h/L (P = 0.002). There was a statistically significant delay overall of non-fasting T max compared with the fasting state T max (P = 0.005). In the multivariate analysis, besides the effect of fasting, C max for females was 20% higher than for males (P = 0.03). Concerning isoniazid, there were significant differences in the C max during non-fasting (median = 3.51 mg/L) compared with fasting (4.54 mg/L). The isoniazid dose received had an effect upon the isoniazid levels (1.26, P = 0.038). In the multivariate analysis, isoniazid exposure during fasting was found to be 14% higher than during non-fasting (CI = 1.02-1.28, P < 0.001). Neither radiological extent of the disease nor consumption of food with drug intake nor pharmacokinetics of rifampicin or isoniazid was associated with a poorer treatment outcome. Conclusions Rifampicin in particular and isoniazid pharmacokinetics were significantly affected by the intake of the drug with food between and within individuals.