Release of organic cations from cationic micelle–clay complexes in slow release herbicide formulations or in column-filters for water purification may pose a problem. In this study we determined the release of octadecyltrimethylammonium (ODTMA) or benzyldimethylhexadecylammonium (BDMHDA) from micelle–clay complexes, and devised a method to retain the released cations by using a column filled with sand/clay mixtures. The adsorption of 5 mM ODTMA or BDMHDA on montmorillonite added at 5 g/l clay was complete. The release was examined at different times and complex concentrations. The percents of release of ODTMA after 24 h were 13.2, 3.8 and 0 for 0.1, 1 and 5 g/L, respectively, whereas for BDMHDA the corresponding values were 31.2, 19.4 and 7.2. Measurements of dissolved organic carbon (DOC) yielded the amount of the released organic cation from micelle–clay complexes in column-filters for water purification after adding a layer (13 cm) of quartz mixed with clay (150:1 w/w ratio) at the bottom of the column. The amount of the released organic cation from columns including the micelle–clay complexes was less (or equal within the experimental error) than the amount released from those including quartz alone or a quartz–clay mixture. Hence, the carbon analysis showed that in the ppm range all the released cations (ODTMA or BDTMA) were retained by the column-filter. The water passing through a 25 cm column filled with quartz mixed with micelle–clay was also passed sequentially through three columns filled with quartz–clay (150:1) and was concentrated 1000-fold. Measurements by GC-MS showed no traces of the organic cations. The conclusion is that the released organic cations from micelle–clay complexes can be completely retained by adding quartz–clay layers.