The efficacy of antiviral drugs and vaccines administered intranasally may depend upon the technique of application. The distribution and time-course of removal of human serum albumin-technetium 99m (HSA-Tc 99m)-instilled intranasally were studied in eleven healthy volunteers using a gamma camera and an anterior sodium iodide scintillation detector. In 100 randomized studies material was delivered as drops in the supine position or as a spray to seated subjects. A significantly higher proportion of 'good' distributions (62 in 73 tests) was obtained with drops compared with spray (1 in 27). The volume administered was varied between 0.10 ml and 0.75 ml and the concentration of HSA was changed from 3 to 30% with no significant effect upon the distribution of time-course of removal; pertechnetate in isotonic saline was distributed and removed in a manner comparable to HSA-Tc 99m. Activity recorded by the detector showed an initial rapid fall associated with removal of most of the material from the nasal cavity, followed by a slower decline associated with the removal of material mainly from the anterior region of the nose. A multidose study confirmed that frequent administration by drops is required to maintain a high level of activity in the nasal cavity. Using this technique it should be possible to correlate measurements of antiviral efficacy and vaccines take-rates with certain characteristics of intranasal applicators; such studies may lead to the design of better devices.