PAF-acether-aggregated rabbit platelets did not respond to a second challenge with the same agonist after their spontaneous disaggregation, but still aggregated in the presence of arachidonic acid (AA). When using 1 nM of PAF-acether in the first stimulation, aggregation in response to a second challenge with the same dose was nil. However, when 0.34 nM was used in the first step, aggregation in response to 0.64 nM PAF-acether was 59 96, as compared to control platelets. By contrast, aggregation of platelets pretreated with 1 nM PAF-acether was 72 % in the presence of 2.8 uM AA. Adding fibrinogen (0.34 mg/ml) before the second stimulation did not modify the desensitization pehnom- enon. Supernatants from platelets desensitized with 1 nM PAF- acether exhibited neither aggregating nor inhibitory activity. PAF-acether-induced desensitization could always be overcome, since aspirin-pretreated platelets first stimulated with 10 nM PAF-acether still aggregated with 100 nM of the same agonist, i.e. when using 100 times the amount which induced maximal aggregation. All our experiments were performed in the presence of ADP scavengers and, except for AA-induced aggregation, aspirin. We tested PAF-acether from 4 different cell origins and the semi-synthetic and synthetic compounds. Platelets were cross- desensitized towards PAF-acether from any source. Totally synthetic PAF-acether bearing aC,, alkyl chain at position 1 of the glycerol moiety desensitized platelets as well as the C.g synthetic analog. Lyso-PAF-acether (i.e. a compound lackifig tne acetyl group at the position 2 of the glycerol) (0.5-10.0 nM) and PAF-acether enantiomer (0.5-5.0 nM) neither aggregated nor desensitized platelets to PAF-acether. These results indicate that 1) PAF-acether from any source exhibit at least an identical active molecular site ; 2) the presence and the stereospecific position of the 2-acetyl group are critical for the interaction of PAF-acether with platelets, a result which could indicate the existence of a platelet acceptor (receptor) for PAF-acether. However, these postulated platelet membrane acceptors were never saturated even using very high amounts of the agonist.