OBJECTIVE To compare measurements of regional myocardial blood flow between color and radioactive microspheres. METHODS Prospective, randomized, controlled. METHODS University research laboratory. METHODS Pigs. METHODS Pigs underwent constriction of the left anterior descending artery, either incremental and then 0 constriction with epinephrine, 0.5 to 3 mu/kg/min (n = 5; "variable") or only 0% and 100% constriction without epinephrine (n = 4, "single"). Radioactive and color microspheres were injected simultaneously. For variable constriction, 5 colors (3 x 10(6)/mL) were tested in random order and, for single, red and yellow (6 x 10(6)/2 mL). RESULTS Measurements of regional endocardial, epicardial, and transmural myocardial blood flow were compared by regression analysis (linear and nonlinear). With radioactive measurements as the point of reference, when regional flow was 50 to 150 mL/min/100 g, correlation was high (r = 0.85), although regression slope was low. With endocardial and epicardial flow between 30 and 100 mL/min/100 g, correlation was close (r = 0.84). Overall nonlinear correlation was higher with single than variable constriction (r = 0.72). When regional flow was less than 100 mL/min/100 g, linear correlation was r = 0.72. When transmural flow measured by color microspheres was less than 25 mL/min/100 g, correlation was high (r = 0.86) but, with endocardial or epicardial flow, low (r = 0.67). When transmural flow was greater than 100 mL/min/100 g, correlation was extremely low (r = 0.1; n = 26 data points). The overall correlations for regional endocardial and epicardial flows were also low, except in the ischemic zone. CONCLUSIONS Color and radioactive measurements correlate well during moderate and ischemic regional myocardial blood flow, ischemic blood flow requiring a higher concentration of color microspheres. A major limitation of using color microspheres is imprecision when flow is greater than 150 mL/min 100 g.