Oral and sublingual isoxsuprine 20 mg were compared with placebo in double blind randomised cross over trial in 7 patients with Raynaud's phenomenon. Skin thermography, plethysmography of the finger tips and direct temperature measurements showed that sublingual isoxsuprine was slightly but consistently superior to the oral form and to placebo; no significant difference was observed between the latter two treatments. It is concluded that sublingual administration of isoxsuprine has advantages over their routes of administration, and that thermography is a suitable technique for qualitative, but less appropriate for quantitative, measurement of drug effects in patients with Raynaud's phenomenon.