Aim: To analyse serum albumin as a prognostic marker in critically ill patients.Material and methods: The present cross sectional study was conducted in the department of Medicine at Chhatrapati Shivaji Subharti Hospital, Subharti Medical College, Meerut, UP. A total of 100 patients were recruited. Detailed history was taken. Patients assessed clinically on day of admission to MICU. Routine investigations like Hb, WBC, platelet count, RFT, LFT, electrolytes and Serum albumin. Serum albumin level was measured on day 1,3,5,7 to MICU. Radiological investigations like X ray, USG, CT Scan were carried according to need without any cost to patient. Serum albumin was compared to parameters like death rate as mortality indicator, morbidity indicators, need for ventilator support and duration of hospital stay. The means and standard deviations of the measurements per group were used for statistical analysis (SPSS 22.00 for windows; SPSS inc, Chicago, USA). Results: All the nonsurvivors have hypoalbuminemia on day three as compared to 48.75% of survivors with statistically significant difference as p<0.05. All the nonsurvivors have hypoalbuminemia on day five as compared to 40% of survivors with statistically significant difference as p<0.05. The total decline in serum albumin in the survivors from admission to day 7 is 0.87 g/dl. In nonsurvivors it is 1.04 g/dl over a period of 7 days. When mean albumin was compared statistically at different intervals, it was found to be statistically significant. Conclusion: Early recognition of patients at high risk of poor outcome can prompt more aggressive management to improve their survival. Serum albumin is a cheap and cost effective and is routinely measured in all critically ill patients. Serial assessment of serum albumin provides useful prognostic information in critically ill patients.