Objective To evaluate the long-term evolution of cardiovascular parameters, lipid metabolism, body composition and bone mass in untreated and treated adult growth hormone deficient patients (AGHD) comparing the differences between the two groups and within each group. Design Seventy-one AGHD-patients were enrolled; 48 received growth hormone (GH) therapy: treated group (TG) and 23 received no GH therapy: control group (CG). In the TG, 22 were childhood-onset (CO) GH-deficient patients, 18–44 years (12 males) and 26 were adult-onset (AO) GH-deficient patients, 27–66 years (10 males). In the CG, 10 patients were AGHD-CO, 20–43 years (8 males) and 13 were AGHD-AO, 25–70 years (8 males). For patients in the TG, GH was administered at a starting dose of 0.1 mg/day, adjusted to maintain IGF-I levels between 0 and 2 SDS for gender and age. At baseline and during the 4th year of replacement therapy or follow-up, the following parameters were evaluated: body mass index, waist circumference, blood glucose, total cholesterol, HDL-cholesterol, LDL-cholesterol, triglycerides, total cholesterol/HDL-cholesterol ratio, systolic and diastolic blood pressure, 2-D echocardiogram with mitral Doppler, bone mineral density (total body, lumbar spine, and femoral neck), bone mineral content (BMC) and body composition. Results In the TG, there was a decrease in diastolic blood pressure (−4.0 ± 1.8 mmHg, p < 0.035) and an increase in blood glucose levels (0.58 ± 0.19 mmol/L, p < 0.025), bone mineral content (0.2 ± 0.0 kg, p < 0.015) and bone mineral density of lumbar spine (0.3 ± 0.1 SDS, p < 0.015) and femoral neck (0.4 ± 0.1 SDS, p < 0.001). All other variables did not show significant changes in any of the two groups. At year 4, changes (delta) differed between patients in the TG and those in the CG with regard to cholesterol levels (TG: −0.27 ± 0.16 mmol/L, CG: 0.34 ± 0.23 mmol/L, p < 0.045), blood glucose (TG: 0.58 ± 0.19 mmol/L, CG: −0.12 ± 0.19 mmol/L, p < 0.025) and BMC (TG: 0.2 ± 0.0 g, CG: 0.0 ± 0.0 g, p < 0.015). An assessment of the changes in variables over time, with and without therapy, considering CO and AO separately, revealed a significant difference in total cholesterol levels during year 4 in CO patients CO (TG: −0.28 ± 0.25 mmol/L and CG: 0.84 ± 0.25 mmol/L, p < 0.015). No differences related to the time of onset of GHD were found in changes in the remaining variables studied. There were no differences related to gender, GHD etiology or the presence of other pituitary hormone deficiencies in the evolution of the parameters analyzed. Conclusions Our 4-year study in GH deficient adults showed significant beneficial effects on some cardiovascular risk parameters and BMC in treated patients. However, there are still unsettled issues regarding long-term benefits and these patients should be carefully monitored.