To the editor: I read with interest the paper by Masaki and Kondo [1] and the subsequent discussion with Terasako [2,3]. Masaki and Kondo demonstrated clearly that halothane and sevoflurane (at concentrations greater than 0.222 and 0.812 mM, respectively) suppress the activity of soluble guanylate cyclase (GC) isolated from rat brain. In our study using isolated rat aorta [4], these anesthetics suppressed cGMP formation induced by nitric oxide (NO) or by an NO-donor drug (sodium nitroprusside, SNP), being consistent with the findings of Jing et al. [5]. In another study of ours using rat cerebellar slices [6], halothane, at concentrations that suppressed the formation of cGMP induced by N-methyl daspartate (NMDA) and d-aspartate, did not alter the effects induced by SNP. Masaki and Kondo suspected that the concentration of SNP we used (0.3 mM) was a supramaximal one [1], but under our experimental conditions, SNP continued to increase cGMP levels in a concentration-dependent manner at least up to 10 mM [7]. Moreover, both d-aspartate (1 mM) and SNP (0.3 mM) increased cGMP to similar levels, and halothane suppressed only the increase induced by daspartate [6]. I do not believe, however, that the results of Masaki and Kondo [1] or Jing et al. [5] conflict with ours [4,6]. Rather, when the findings of the above four papers are considered together, it is apparent that halothane suppresses GC activity, but that some site or mechanism between the glutamate receptor and NO formation is more susceptible to halothane than GC activity itself. The locus of action is possibly one of the mechanisms coupled to the glutamate receptor.