In a prolonged field trial a 4% chlorhexidine digluconate detergent scrub (Hibiscrub(R)), that had earlier proved to be an effective hand disinfectant, was studied in hospital wards. Finger tips were found to harbour more bacteria than the hand dorsum and the samples collected from them yielded more information on the bacteriological and dermatological effects of hand disinfectants in practice.In wards with a relatively low hand-washing frequency (less than 20 times in 8 hours) the bacteriological results resembled those obtained by in-use tests with volunteers. In the neonatal unit where the hand washing frequency was remarkably high, even occasionally over 100 times/8 h shift, an increase in the bacterial colony counts of the majority of the staff was recorded both before and after hand washing already after using the preparation for 1 week. Age, occupation and hand-washing frequency all correlated with the bacteriological results. Twenty-seven out of 37 persons complained of side effects such as wounds of finger tips and redness or heavy drying of the skin. Wounds, particularly on finger tips, resulted in the failure of disinfection. An increase in bacterial counts was sometimes noted without any dermatological or subjective changes. Drying of the skin was complained of less often when no increase in skin bacteria occurred.After the changeover of washing practice to a detergent followed by a rinse with spirit solution containing chlorhexidine and glycerol a decrease was recorded in the bacterial counts. It is concluded that more attention should be paid to long-term testing of hand washing and disinfection methods to ensure optimum final results in practice. It is obvious that the knowledge obtained from short time in-use testing cannot be applied to all conditions of use.