Rhinovirus was recovered from the fingers of 16 of 38 volunteers and others, who were swabbed during the acute stages of their colds. Very low titres of virus were also recovered from 6 of 40 objects recently handled by infected volunteers, but not from the fingers of 18 non-infected subjects whose flat-mates were shedding virus. When rhinovirus from nasal secretions was dried on skin or other surfaces during laboratory experiments, approximately 40-99% of infectivity was lost. Virus could be transferred from surface to surface by rubbing, the transfer being more efficient if it was carried out while the inoculum was still damp. Volunteers could infect themselves if a moderately heavy dose (88 TCD50) of virus was inoculated on the finger and then rubbed into the conjunctiva or nostril, especially if the inoculum was still damp. From estimates of virus titres in nasal washings and on fingers, and of amounts transferred by rubbing, it was concluded that apread of colds is unlikely to occur via objects contaminated by the hands of the virus-shedder, but that a receipient might pick up enough virus on his fingers by direct contact with heavily infected skin or secretions to constitute a risk of self-inoculation via the conjunctiva or nostril.