We describe a method for measuring 224Ra and 228Th activities in coastal sediments based on a delayed-coincidence counting system (the RaDeCC system). Milli-Q water is added to bulk sediment to form a slurry. 224Ra in interstitial water is co-precipitated by MnO2 suspension. The MnO2 suspension and the sediment with absorbed 224Ra and 228Th are filtered onto a 142-mm 0.7 ¦Ìm (nominal pore size) GFF filter. The filter is placed onto a sample holder specified for sediment samples and counted in a RaDeCC system. 224Ra and 228Th activities are calculated from two measurements that are conducted within 6-12 h and in 8-10 days after sample collection. The RaDeCC system is calibrated with a 232U-228Th standard using the method of standard addition. The reproducibility and the overall accuracy of 224Ra and 228Th measurements based on this method are estimated to be ¡À 5 % and ¡À 5-7 % , respectively. We have applied this method to a coastal sediment and observed a significant deficit of 224Ra with respect to 228Th in the upper 3-4 cm.