Fluoride acquisition by human enamel from topical application of either fluoride solutions or self-gelling fluoride systems with tetraethoxysilane (TES) was studied in vitro. Also studied was fluoride deposition from self-gelling systems incorporating certain positively charged polymers (Jaguar C-15 or DEAE dextran) to reduce the rate of fluoride release from the gels. The systems tested were all 0.65 M with respect to fluoride and of neutral pH. Sodium fluoride with 2% cetylpyridinium chloride (CPC) and ammonium fluoride with 1% sodium lauryl sulfate (SLS) were tested either as solutions or as self-gelling preparations with 20% TES. Additional studies involved gels with (1) sodium fluoride, 0.5 M sodium phosphate and 1% SLS; (2) sodium fluoride with 5% Jaguar C-15 and 1% CPC; (3) ammonium fluoride, 0.5 M ammonium phosphate and 1% SLS, and (4) ammonium fluoride with 5% DEAE dextran and 1% Tergitol 15-S-12. The findings showed that self-gelling systems deliver more fluoride to the enamel than their solution counterparts. The amount of firmly bound fluoride resulting from topical interaction was also greater from the self-gelling sodium fluoride system than from the application of an analogous sodium fluoride solution throughout the outer 20 μm of the enamel. The ammonium fluoride self-gelling system was only superior to the solution in this respect at the very enamel surface (i.e. depth of 2.5 μm). Incorporation of positively charged polymers in the gel systems resulted in decreased deposition of total fluoride and also of firmly bound fluorid