Background: The skin sympathetic nerve activity (SKNA) is a new method to measure sympathetic nerve activity by using conventional ECG electrodes. We developed a novel approach to analyze the complexity of SKNA time series under different time scales and showed its prognostic significance in patients receiving critical care. Methods: This study measured SKNA in patients admitted to an intensive care unit (ICU). Each recording is 10-minute long with 10000Hz sampling rate. Multi-scale fluctuation analysis (MSFA) was developed to quantify the variation within each time scale after removing the linear trend. The prognostic value of SKNA was combined with traditional prognostics scoring system to improve the predictive values. Results: 155 patients were recruited. After 30 and 90 days, 30 and 48 patients expired. MSFA was significantly higher in survival group than mortality group for 30-day (0.487 ± 0.185 vs 0.401 ± 0.045, p = 0.018) and 90-day (0.499 ± 0.196 vs 0.414 ± 0.061, p = 0.001) follow-up. Sequential Organ Failure Assessment (SOFA) score was significantly lower in the survival group compared to the expired group for 30-day and 90-day (4.1 ± 2.9 vs. 5.5 ± 4.1, p = 0.032 and 3.9 ± 3.0 vs. 5.4 ± 3.5, p = 0.012). The Kaplan–Meier survival analysis showed MSFA lower than 0.401 (log-rank test:4.96, p = 0.03) or with SOFA score lower than 5 (log-rank test:5.49, p = 0.019) have a significantly higher mortality rate. A multivariate Cox regression model showed that the MSFA is an independent predictor for 30-day mortality (HR = 2.35, 1.08–5.09, p = 0.031) and 90-day mortality (HR = 1.96, 1.08–3.58, p = 0.027). Conclusion: MSFA was a significant prognostic predictor for critically ill patients. MSFA adding to SOFA score could help improve risk prediction.