The multicenter OP-RISK study, developed during 1994-96, was aimed at: 1) investigating early (28 days) death rates following aortocoronary bypass surgery among patients recruited from four Centers representing geographical distribution in Italy; 2) defining possible risk factors for early mortality, also comparing these factors with those reported in previous studies. Average values are reported and compared of 65 variables (36 preoperative, 10 operative and 19 postoperative) out of 984 patients subdivided into alive (n = 940) or dead (n = 44, 4.47%) at 28 days (155 +/- 174 hours, interval between 12 and 576 hours) postoperatively. Causes of death were cardiac in 37 (77%), pulmonary in 3 (0.7%), vascular in 2 (0.5%) and infective in 2(0.5%) patients, respectively. During the study a total of 1126 patients were operated upon in the collaborative Centers with the diagnosis of coronary artery disease and 51 deaths were reported officially in-hospital (4.53%). Therefore, OP-RISK data represent 87% of overall patients and a superposable death rate. The potential role as risk factors of early mortality was assessed univariately for 17 preoperative, 5 operative (in 3 cases for the first time) and 5 postoperative factors. In general, it was confirmed that factors defining left ventricular function are sensitive predictors of mortality. In OP-RISK we were able to show, in addition, that tachycardia (> 130 b/min) at induction of anesthesia, and total time of anesthesia, cardiopulmonary bypass and aortic cross clamping may be significant factors among operative variables as might be among postoperative ones several arrhythmia types or a lower rate in antithrombotic therapy with aspirin at 6-12 hours postoperatively. The protective role of bypass surgery performed with at least 1 arterial segment was also ascertained. Most of these potential factors were significantly related to outcome (either directly or inversely) as were among them, as seen in a subsample (65%) of 639 patients in whom a correlation matrix was performed among 16 factors selected on the basis of the common denominator principle. Our results suggest that it is possible to collect in a multicenter experience univariate predictors of early mortality following aortocoronary bypass surgery in Italy, which are not different from those reported from previous studies performed abroad. Operative indicators may also have predictive capabilities. The effort may be worthwhile and demands further cooperative studies to be undertaken, aimed at obtaining nationwide coefficients of risk along with representative average values of factors that soon might emerge once multivariate statistics will be performed on this material.