A randomized controlled study that evaluated a recall system and patient education material by mail in 178 asymptomatic female family practice patients aged 50 to 69 years showed no effect on the proportion of patients who had cancer screening tests (P = .20) and a significant adverse effect on the mean number of tests performed (P = .05) after 4 months. In a subgroup of previous compliers (those who had one or more tests 12 months before the study), however, there was a lower proportion of patients receiving one or more tests (P = .019) with a lower mean number of tests (P = .007) than previous compliers in the control group. Recall strategies for cancer screening tests need to be more extensively studied in the United States before they are routinely adopted in family practice.