An attempt to find a reliable way to measure contraceptive continuation among teenagers was made in a study of 445 adolescents under age 18 who made an initial visit to one of nine federally funded family planning clinics in the Philadelphia area in 1980 and 1981. The participants were interviewed at their initial visit and at six and 15 months to collect information on background characteristics, pregnancy history and contraceptive use. A cross-check of the information on contraceptive use reported during the study showed that 38 percent of the respondents who reported that they had been continuous users during the 15 months supplied information at other points indicating that they had not always used contraceptives during that period. In an attempt to examine the effects of adolescent reporting errors on measures of contraceptive use, the investigators constructed three alternative measures of contraceptive continuation. The simplest measure assessed current use at 15 months. A second measure accepted only reports of continuous use at 15 months and ignored inconsistent data. The third measure corrected for inconsistent data. Applying these measures to the study population produced proportions of contraceptive continuation ranging from a low of 43 percent when the most stringent measure was used to a high of 82 percent for the current-use category. Whatever the measure used, much of the discontinuation appears to have occurred in the first three months after the initial visit to a family planning clinic.(ABSTRACT TRUNCATED AT 250 WORDS)