Background The complexity and cost of measuring weight-for-height make it unsuitable for use by community-based volunteers. This has led community therapeutic care programs to adopt a two-stage screening and admission procedure in which mid-upper-arm circumference (MUAC) is used for referral and weight-for-height is used for admission. Such a procedure results in many individuals being referred for care on the basis of MUAC but subsequently being refused treatment because they do not meet the weight-for-height admission criterion. This “problem of rejected referrals” has proved to be a major barrier to program uptake. Objective To systematically review methods to detect cases of severely malnourished children in the community for their admission into community-based therapeutic care programs. Methods Clinical and anthropometric methods for case detection of severely malnourished children in the community were reviewed with regard to their ability to reflect both mortality risk and nutritional status. Results MUAC, with the addition of the presence of bipedal edema, was found to be the indicator best suited to screening and case detection of malnutrition in the community. The case definition “MUAC < 110 mm OR the presence of bipedal edema,” with MUAC measured by a color-banded strap, is suitable for screening and case detection of malnutrition in the community for children aged between 6 and 59 months. Monitoring and discharge criteria were also reviewed. Conclusions There is no compelling evidence to support a move away from using weight in combination with clinical criteria for monitoring and discharge.