The effects of systemically-administered human alpha calcitonin gene-related peptide (h.alpha CGRP), substance P and the selective neurokinin receptor agonists, GR73632 (NK1) and GR64349 (NK2) on cerebral blood flow (CBF) were studied in anaesthetized guinea-pigs using a laser-Doppler flowmeter. h.alpha CGRP (0.1 and 0.3 nmol/kg), substance P (0.03-1.0 nmol/kg), GR73632 (0.03-0.3 nmol/kg) and GR64349 (0.3 nmol/kg) each, following intra-carotid artery injection, reduced transiently (< 5 min) blood pressure and CBF. GR73632 (0.1 and 0.3 nmol/kg) and GR64349 (0.3 nmol/kg), but not h.alpha CGRP (0.01-0.3 nmol/kg) or substance P(0.01-1.0 nmol/kg), then produced a more prolonged increase in CBF, the peak effect occurring 10-15 min after injection. It is likely that this increase in CBF was due to their bronchoconstrictor activity, rather than a direct effect on the cerebrovasculature; arterial PaCO2 levels were increased and PaO2 decreased by both compounds. Following pretreatment with urea (5 M) to disrupt the blood brain barrier, h.alpha CGRP (0.1 nmol/kg) produced a significant increase in CBF (13 +/- 4%), implying that access to its receptors on the cerebrovascular smooth muscle had been achieved. Substance P (0.1 nmol/kg) remained inactive. The study has demonstrated that compounds acting on neuropeptide receptors have little direct influence on CBF following systemic administration. CGRP requires access to its receptors on the cerebrovascular smooth muscle, while selective NK1 and NK2 receptor agonists increase CBF, probably indirectly via their bronchoconstrictor activity. The lack of effect of substance P may be due to its rapid breakdown by peptidases, a property not shared by the selective neurokinin agonists.