Child sexual abuse research presents practical, methodological and statistical challenges. This paper examines some recent literature about the incidence and effects of child sexual abuse. Researchers who use epidemiological studies to attempt to ascertain the prevalence of different forms of abuse should be aware of the different populations studied, different methodologies used, diverse definitions of sexual abuse and practical limitations on participants' memories. Common problems in retrospective and prospective studies into the effects of child sexual abuse include the use of atypical populations, inappropriate statistical analyses, the use of unvalidated psychometric instruments and limited sources of information on child functioning. Attention must also be paid to potentially confounding demographic and family variables. A thorough understanding of, and attention to, these problem areas in child sexual abuse research will give a clearer picture of the prevalence and effects of abuse, thus indicating the most appropriate forms of prevention and treatment.