This paper reports on a study that compared the clinical performance of NORPLANT implants in two separate but similar cohorts: 200 NORPLANT users from 1975 through 1978, when the method was first introduced in the clinic, and 212 users in the same population from March 1982 through December 1983. Two different cohorts of Copper T IUD users were included as controls. Continuation rates, at one and two years of use, were significantly higher in the second group of NORPLANT users (88.1 vs. 60.5 and 76.5 vs. 40.8, respectively). Statistically significant differences between the NORPLANT cohorts were observed in discontinuation due to bleeding, amenorrhea, other medical causes, total medical causes, and personal reasons. No differences in clinical performance of the IUD were observed between the two control groups. The higher continuation rate in the second group of implant users is attributed to greater confidence and experience among the clinic staff between 1982 and 1983, and greater knowledge of and familiarity with the method among those in the target population. The study also shows that the first clinical trials of a new contraceptive method may not provide an accurate picture of that method's ultimate performance.