An intercomparison of exposure secondary standards for soft roentgen rays in use in the Nordic countries is reported. One of the standards was used as a reference standard against which all the other standards were calibrated. The statistical uncertainty in the calibrations was estimated to be 0.8 per cent (at the 95% confidence level) and the maximum systematic uncertainty in the intercomparison was estimated to be 0.8 per cent. The observed differences, at most about one per cent, were not considered significant. It was concluded that for this type of chambers a period of about five years is acceptable between recalibrations against a primary standard unless stability checks indicate otherwise.