OBJECTIVE Minimally invasive direct coronary artery bypass (MIDCAB) provides many anesthetic challenges including monitoring, managing myocardial ischemia, and pain control. The objective was to evaluate the monitoring requirements and the potential benefits of preischemic conditioning and intrathecal morphine sulfate in MIDCAB patients. METHODS This review was retrospective and unrandomized and was conducted at Allegheny University Hospitals, Allegheny General, Pittsburgh, PA. METHODS Sixty-four patients with single coronary artery lesions (> 70% obstruction) underwent attempted MIDCAB during a 1-year period between November 1995 and November 1996. Seven patients required conversion to conventional coronary artery bypass grafting (CABG) with cardiopulmonary bypass (CPB) and two patients required extended thoracotomy incisions. This report describes the remaining 55 patients who underwent MIDCAB. METHODS Some of the MIDCAB patients received intrathecal morphine before anesthetic induction. Ischemic preconditioning was assessed in a subset of patients. RESULTS MIDCAB was performed in 55 of 64 patients. Transesophageal echocardiography (TEE) was used in all patients and a pulmonary artery catheter was used in 43% of patients. Esmolol was used in 25% of patients to reduce motion of the left ventricle (LV) during the left internal mammary artery (LIMA)-LAD anastomosis, but was used less often as the surgeons adapted to the use of a retractor that stabilized the ventricular wall adjacent to the site of the LIMA-LAD anastomosis. LAD occlusion caused reversible, regional systolic dysfunction by TEE in the anterior and apical LV segments. During LAD occlusion, nitroglycerin was used in 61% of patients and phenylephrine in 24%. Ischemic preconditioning did not prevent increases in systemic or pulmonary artery pressures during LAD occlusion. Most (85%) patients were extubated in the operating room. Intrathecal morphine decreased postoperative analgesic requirements. The mean hospital length of stay (LOS) was 4.0 +/- 1.7 days (range, 1 to 10 days). CONCLUSIONS MIDCAB may reduce hospital LOS for patients with single vessel coronary artery lesions when compared with median sternotomy with a LIMA-LAD graft performed on cardiopulmonary bypass. Pharmacologic heart rate control during the LIMA-LAD anastomosis is not critical with the use of a surgical retractor which diminishes ventricular motion. A single 5-minute test LAD occlusion did not protect against subsequent regional ischemic dysfunction in our subset of patients with normal baseline function.