The potential of the near infrared spectroscopy (NIRS) technique for the analysis of red paprika for aflatoxin B 1, ochratoxin A and total aflatoxins is explored. As a reference, the results from a chromatographic method with fluorescence detection (HPLC-FD) following an immunoaffinity cleanup (IAC) were employed. For the NIRS measurement, a remote reflectance fibre-optic probe was applied directly onto the samples of paprika. There was no need for pre-treatment or manipulation of the sample. The modified partial least squares (MPLS) algorithm was employed as a regression method. The multiple correlation coefficients (RSQ) and the prediction corrected standard errors (SEP(C)) were respectively 0.955 and 0.2 μg kg −1, 0.853 and 2.3 μg kg −1, 0.938 and 0.3 μg kg −1 for aflatoxin B 1, ochratoxin A and total aflatoxins, respectively. The capacity for prediction of the developed model measured as ratio performance deviation (RPD) for aflatoxin B 1 (5.2), ochratoxin A (2.8) and total aflatoxins (4.4) indicate that NIRS technique using a fibre-optic probe offers an alternative for the determination of these three parameters in paprika, with an advantageously lower cost and higher speed as compared with the chemical method. Content of aflatoxin B 1 and total aflatoxins are the parameters currently employed by the food regulations to limit the levels of the four aflatoxins in many foodstuffs. In addition, aflatoxin B 1 itself is an excellent indicator for aflatoxins’ contamination since it is always the most abundant and toxic.