In this study, two method intercomparisons were performed. One thermal and two optical methods for the measurement of black carbon (BC) were applied to laboratory generated aerosols containing only BC. For the optical measurements, an aethalometer (Hansen et al., 1984. Science of Total Environment 36, 191-196) and an integrating sphere technique (Hitzenberger et al., 1996b. Journal of Geophysical Research 101, D14, 19 601-19 606) were used. The thermal method was described by Cachier et al. (1989a. Tellus 41B, 379-390). In an additional comparison, the integrating sphere was compared to a thermal optical technique (Birch and Cary, 1996. Aerosol Science Technology 25, 221-241) on ambient aerosol samples. The absorption coefficients were obtained from transmission measurements on filter samples for both the aethalometer and the integrating sphere. The BC mass concentration for the aethalometer was derived from this absorption measurement. The BC mass concentration for the integrating sphere, however, was obtained using an independent calibration curve. The agreement between the absorption coefficient sigma(a) obtained for the BC test aerosol on parallel filters with the aethalometer and the integrating sphere was satisfactory. The slope of the regression lines depended on filter type. A comparison between BC mass concentrations, however, showed that the aethalometer values were only 23% of those obtained by the integrating sphere technique indicating that for pure BC aerosols, the standard aethalometer calibration should not be used. Compared to the thermal method, the integrating sphere gave an overestimation of the BC mass concentrations by 21%. For the ambient samples, the integrating sphere and the thermal optical methods for BC mass concentration determination showed agreement within 5% of the 1 : 1 line, although the data were not so well correlated. (C) 1999 Elsevier Science Ltd. All rights reserved.