Growth hormone (GH) pituitary tumors are almost always benign adenomas, yet are associated with significant morbidity and mortality. Surgical and medical responses of GH tumors are often incomplete, and therefore predictors of residual or recurrent disease are needed. Clinical features, including patient gender, age or size of adenoma, have proven to be unreliable predictors of recurrence. Differing clinical behavior between the two GH tumor subtypes, sparsely granulated (SG) versus densely granulated (DG), has been reported, but has not been used routinely in clinical management. SG tumors are more common in younger patients (<50 years), and are usually larger tumors. SG tumors have been reported to be less responsive to somatostatin analogs (SSA) than DG tumors. The mechanisms underlying these potential differences in tumor behavior, however, are poorly defined. Subsets (up to 50 %) of DG adenomas harbor a gsp mutation that can activate cAMP that provides a theoretical intracellular target for somatostatin therapy. In contrast, some SG tumors have reduced somatostatin receptor expression and mutations in the extracellular domain of the GH receptor that may contribute to SSA resistance. While DG versus SG growth hormone adenomas are readily distinguished by immunohistochemistry, other less common GH adenoma variants still require electron microscopy (EM) for confident subclassification. Whether these less common variants possess unique clinical features is unknown. Research is needed to identify clinically relevant biomarkers of GH pituitary tumors that predict risk of recurrence and response to medical therapy.