Immunohistochemistry is commonly performed on tumour specimen obtained during transsphenoidal surgery, but its predictive value for clinical outcome is largely unknown. The aim of this study was to compare clinical and biochemical outcome characteristics after surgery with histological tumour properties. This was achieved by matching data from the German Acromegaly Register with those of the Pituitary Tumour Registry of the German pituitary working group. From 285 out of 1543 acromegalic patients of the German Acromegaly Register (145 f,140 m), data on morphological properties analyzed by a single pathologist (W.S) in the department of pathology, Marienkrankenhaus, Hamburg were available. Using immunohisto-chemistry, the density of cytoplasmatic granules, pattern of hormone expression and mitotic activity (Ki67) were analyzed. Tumours were stratified according to growth hormone (GH) and prolactin expression and Ki67 index. Clinical and biochemical parameters predicting disease outcome such as post-surgical GH and IGF-1 were analyzed. Control of acromegaly was defined as random GH <2.5 μg/l and normal IGF-1. Results are presented as range, mean and SEM. Before surgery, GH and IGF-1 concentration did not differ between patients with sparsely (n=93) and densely granulated (n=145) adenomas. However, after transsphenoidal surgery, patients with densely granulated adenomas had significantly higher GH (0100, 5.4±1.14 vs 057, 2.98±0.917 μg/l, P=0.03) and IGF-1 (941963.522±14.81, vs.121002.456±36.38 ng/ml, P=0.006) concentrations compared to sparsely granulated adenomas. These patients had a lower rate of bio-chemical control (31% vs 54%, P=0.01). Co-expression of prolactin was found in 14% of adenomas. This was associated with higher postsurgical GH and IGF-1 (GH 10.5±8.31 vs 3.3±1.23μg/l, IGF-1 437±149.01 vs 348±27.3 ng/ml) compared to tumours not expressing prolactin. Ki67 staining (Ki67 index <1% vs >1%) did not have impact on clinical and biochemical variables (P=n.s.). The granulation density of GH producing adenomas is a useful parameter predicting patients biochemical outcome in acromegaly.