The purpose of this study is to examine the relationship between time in situ and the frequency of thrombophlebitis. Nurses' care and handling when using a peripheral intravenous (p.i.v.) line was also studied. The methods used were observation, interview, and document analysis after intervention, with respect to cannula time in situ. The study included 120 patients divided in two groups: (1) experimental, time in situ less than or equal to 24 hours, and (2) control, using current/daily routines from the wards. The frequency of thrombophlebitis after a p.i.v. was inserted was significantly higher and showed more troublesome and prolonged picture complications in the control group than in the experimental group. Care and handling suffered with increased number of days in situ. The nurses' documentation of inserted or removed cannula was incomplete in most cases, and notices about the insertion area were nearly nonexistent. It was determined that a short time in situ--when using a p.i.v.--is an important factor in preventing complications.