The agonist dissociation constant for 5-hydroxytryptamine (5-HT) was estimated in the guinea-pig isolated trachea by the method of receptor inactivation. The value obtained (pKA = 6.45) was significantly lower than estimates previously obtained in the rabbit aorta and rat jugular vein, although all three tissues are supposed to contain the same 5-HT2 class of receptor. 2. The antagonist dissociation constant for alpha,alpha-dimethyltryptamine was also estimated in the guinea-pig trachea. The pKB value (5.43) was not significantly different from previous estimates in the rabbit aorta and rat jugular vein, consistent with receptor homogeneity between the three tissues. 3. The effect-time profiles corresponding to individual 5-HT applications were more transient in the guinea-pig trachea than in the rabbit aorta. This difference could be accounted for using a simple model of acute receptor desensitization (Leff, 1986), assuming that the conversion of active agonist-receptor complexes into inactive ones was faster in the guinea-pig trachea than in the rabbit aorta. 4. Computer simulation of the desensitization model showed that the discrepancy of pKA estimates for 5-HT between the rabbit aorta and guinea-pig trachea could also be explained using the same rate constant difference that accounted for the difference in effect-time profiles. This analysis indicated that the estimate made in the trachea was erroneously low, whereas that made in the aorta was concluded to be correct. 5. The apparent association between transience of response and pKA estimates is discussed with particular attention to the reliability of agonist affinity estimates in receptor classification.