The early treatment of hyperlipidemia in hospitalized patients confers potential benefit, yet total cholesterol is known to vary with acute illness, often delaying treatment decisions. A prospective study was conducted of 61 patients (mean age 57 years; 49% women) admitted to an acute care community hospital with various diagnoses with random nonfasting lipid profile measurements at admission, followed by second fasting lipid profile measurements on day 3 of hospitalization or upon discharge (whichever occurred first), and final fasting lipid profile measurements 4 weeks after discharge. All individual values of the lipid profile decreases at discharge, whereas the ratios of total cholesterol to high-density lipoprotein (HDL) and of low-density lipoprotein (LDL) to HDL did not change significantly. The 95% confidence interval around the total cholesterol/HDL ratio for each patient was within the same National Cholesterol Education Program Adult Treatment Panel III treatment recommendation category 42.6% of the time, whereas corresponding intervals for total cholesterol and LDL were within a single treatment category only 6.6% of the time. The total cholesterol/HDL ratio was significantly more consistent with regard to treatment implications than LDL or total cholesterol (p <0.0001). In conclusion, serum lipid values vary significantly during and after a hospital stay, whereas the ratio of total cholesterol to HDL remains relatively stable. This ratio may therefore serve as a more reliable basis for early treatment decisions in dyslipidemia.