In treatment-resistant patients with acromegaly, pharmacotherapy with pegvisomant (Somavert) is a highly effective option. However, safety concerns have been raised related to a potential increase in tumor size during long-term pegvisomant treatment. Therefore, neuroradiological monitoring of tumor extension and volume was performed in the German Pegvisomant Observational Study, which covers 87% of patients treated with pegvisomant in Germany. As of 15 July 2007, a total of 307 patients (156 males and 151 females) had been included in the study and were on pegvisomant therapy for an average of 86.7 weeks. Median and mean doses of pegvisomant were 15 and 16.6 mg/day respectively. Out of these 307 patients, 18 were reported to have tumor-size increases as adverse events. From these 18 patients, all available serial magnetic resonance images were collected. Identical or similar sequences were chosen and the region of interest was magnified and compared across time after the best possible fit had been achieved by size and gray-scale correction. All available images were carefully re-evaluated according to this method. In 10 out of the 18 patients, there was no evidence of tumor-size increase, when the pre-treatment scans were compared with the most recent follow-up investigations. In two out of the remaining eight patients, there was a rebound effect observed after withdrawal of somatostatin analog treatment, but no further progression. In another three out of the eight patients, tumor-size increase had already been documented before pegvisomant treatment was commenced, during preceding somatostatin analog treatment and continued therapy. In the last three patients, tumor progression after the start of pegvisomant treatment was confirmed. All three patients had undergone pituitary surgery as primary treatment, but had not been pre-treated with radiotherapy. In all three cases, the tumor increase was not considered clinically significant and the investigators decided to continue pegvisomant treatment. In conclusion, in this large group of pegvisomant-treated patients, tumor progression was rare. It was reported in between 2 and 3% of patients treated, and did not exceed the expected rate in patients with acromegaly not treated with pegvisomant. In over one-half of patients, reports of tumor increase could not be confirmed by re-evaluation. This was mostly due to non-identical gantry projections. Misjudgements mainly occurred when only images from two individual investigations, rather than the entire series of scans, were compared. Thus, we recommend a careful serial evaluation of all available images to avoid misinterpretations and erroneous alerts. As from this presently largest database of acromegalic patients treated with pegvisomant, tumor-growth rate appears not to be different from patients on other treatment modalities. Although these data are reassuring with regard to the concern of somatotroph adenoma growth under peripheral GH receptor blockade, further study is required.