The value of follow-up surveillance for patients with cutaneous melanoma remains uncertain. In this prospective study the frequency of detection of first melanoma recurrence (FMR) by patient or doctor was analyzed to assist in the future design of evidence-based follow-up guidelines. Patients who had a recurrence of a previously treated American Joint Committee on Cancer (AJCC) stage I-III primary melanoma (PM) were interviewed to ascertain how their PM and FMR were detected. Factors predicting the detection of PM and FMR were analyzed. The study group comprised 211 patients. In 168 patients, information on detection of their PM was available; 102 PMs (61%) were detected by the patient and 18 (11%) by their partner. Higher AJCC stage, visible location for the patient, and female sex were independent predictive factors for patient-detected PM (P = .03, .002, and .02 respectively). The FMR type was local in 28 (13%), in transit in 35 (17%), in regional lymph nodes in 97 (46%), and distant in 51 (24%). Seventy-three percent of all FMRs were detected by the patient. The presence of a symptom was the only independent predictor of a patient-detected FMR (P < .0001). There was no statistically significant survival difference between the patient-detected and doctor-detected FMRs. Three-quarters of FMRs were detected by patients or their partners, and it should be possible to improve this rate even further by better education. More frequent follow-up visits are thus unlikely to be valuable. Reductions in follow-up frequency may therefore be safe and economically responsible.