The ability to effectively define disease status in ovarian cancer after initial therapy or to selectively screen high-risk populations remains a major challenge for in vivo monoclonal antibody (mAb)-targeted approaches. Antitumour murine mAbs (HMFG1, HMFG2, H317, and H17E2) and the reshaped human antibody Hu2PLAP (against placental alkaline phosphatase; PLAP), labelled with indium-111 and iodine-123, were evaluated for their ability to localise ovarian tumours in sequential studies of our group. Thirty patients with ovarian cancer, aged 40-78 years (median 60 years) were studied with HMFG1/G2: 11, and H317/H17E2: 19 murine mAbs. Six patients with ovarian cancer aged between 36 and 65 years (median 49 years) were studied with the reshaped human Hu2PLAP mAb (5 patients) or the murine H17E2 mAb (2 patients) labelled with 111In via a new macrocyclic chelating agent (DOTA). One of these was imaged twice, with H17E2- and Hu2PLAP-DOTA-111In, respectively. In 20 out of 22 patients with radiologically measurable ovarian cancer, the presence of tumour was confirmed by the murine mAb scan and correlated well with the findings of conventional radiology diagnostic methods. One of these patients with a negative H17E2 scan and a large abdominal mass at laparotomy was found to have a PLAP-negative tumour on immunohistochemistry. Additionally, the antibody scan revealed the presence of active disease, confirmed at laparotomy/laparoscopy, in 6 out of 8 patients considered to be in clinical complete remission. Best images were obtained at 24 and 48 h after the 123I and 111In mAbs, respectively. Successful imaging with the reshaped human antibody, Hu2PLAP, was seen in 2 patients with PLAP-positive tumours. Antibodies to DOTA developed in 2 patients. In conclusion, immunolocalisation of ovarian tumours is feasible with both murine and reshaped human mAbs. The sensitivity and specificity of the method appear very high in this pilot study, and in view of the absence of toxicity, the diagnostic contribution of this approach should be evaluated prospectively. Given the low number of patients without surgically detectable disease in the present study, future investigations should include more patients with no evidence of disease in order to provide more meaningful estimates of specificity.