Unintended pregnancies account for about half of all pregnancies in the United States and, in 1995, numbered nearly 3 million pregnancies. They pose appreciable medical, emotional, social and financial costs on women, their families and society. The US is not attaining national goals to decrease unintended pregnancies, and little is known about effective means for reducing unintended pregnancy rates in adults or adolescents.To examine the evidence about the effectiveness, benefits and harms of counseling in a clinical setting to prevent unintended pregnancy in adults and adolescents and to use the evidence to propose a research agenda.We identified English-language articles from comprehensive searches of the MEDLINE, CINAHL, PsychLit and other databases from 1985 through May 2000; the main clinical search terms included pregnancy (mistimed, unintended, unplanned, unwanted), family planning, contraceptive behavior, counseling, sex counseling, and knowledge, attitudes and behavior. We also used published systematic reviews, hand searching of relevant articles, the second Guide to Clinical Preventive Services and extensive peer review to identify important articles not otherwise found and to assure completeness. Of 673 abstracts examined, we retained 354 for full article review; of these, we used 74 for the systematic evidence review and abstracted data from 13 articles for evidence tables. Four studies addressed the effectiveness of counseling in a clinical setting in changing knowledge, skills and attitudes about contraception and pregnancy; all had poor internal validity and generalizability and collectively did not provide definitive guidance about effective counseling strategies. Nine studies (three in teenage populations) addressed the relationship of knowledge on contraceptive use and adherence. Knowledge of correct contraceptive methods may be positively associated with appropriate use, but reservations about the method itself, partner support of the method, and women's beliefs about their own fertility are important determinants of method adherence that may attenuate the knowledge effect. Many factors influence contraceptive use and adherence; among them are age, marital status, ambivalence about becoming pregnant, attitudes of partner, side effects, satisfaction with provider and costs; however, the impact of such factors may not be consistent across populations defined by cultural, age or other factors. The studies themselves differed materially in outcome variables, populations and methodologies and did not yield a body of work that can reliably identify specific influences on contraceptive use and adherence. No literature reports on harms of counseling or on the costs or cost-effectiveness of different approaches to counseling about unintended conceptions in the primary care setting. Virtually no experimental or observational literature reliably answers questions about the effectiveness of counseling in the clinical setting to reduce rates of unintended (unwanted, mistimed) pregnancies in this country. Existing studies suffer from appreciable threats to internal validity and loss to follow-up and are extremely heterogeneous in terms of populations studied and outcomes measured. The quality of the existing research does not provide strong guidance for recommendations about clinical practice but does suggest directions for future investigations. Numerous issues warrant rigorous investigation.