Prospective memory problems are common in patients with brain injury, but appropriate measures are limited. The reliability and validity of the newly designed Royal Prince Alfred Prospective Memory Test (RPA-ProMem), which has three alternate versions, was investigated in 20 healthy volunteers and 20 neurological patients with everyday prospective memory problems. The RPA-ProMem was found to be easy to score reliably (inter-rater reliability = .90) and its three versions were well matched (delayed alternate-form reliability = .71). Test validity and sensitivity to patient deficits were also supported. This new measure of prospective memory should be particularly useful in situations that require repeated assessments, such as evaluation of rehabilitation efforts.