Objective: The aim of this study is to evaluate the effectiveness and efficiency of causal links between various error causesin human error analysis. Background: As finding root causes of human error in safety-critical systems is often a cognitivelydemanding and time-consuming task, it is particularly necessary to develop a method for improving both the quality andefficiency of the task. Although a few methods such as CREAM have suggested causal linking between error causes as ameans to enhance the quality and efficiency of human error analysis, no published research to date has evaluated theperformance of the causal links. Method: The performance of the CREAM links between error causes were evaluated with80 railway accident investigation reports from the UK. From each report, errorneous actions of operators were derived, andfor each error, candidate causes were found by following the predefined links. Two measures, coverage and selectivity, wereused to evaluate the effectiveness and efficiency of the links, respectively. Results: On average, 96% of error causes actuallyincluded in the accident reports were found by following the causal links, and among the total of 121 possible error causes,the number of error causes to be examined further was reduced to one-tenth on average. As an additional result of this work,frequent error causes and frequently used links are provided. Conclusion: This result implies that the predefined causal linksbetween error causes can significantly reduce the time and effort required to find the multiple levels of error causes and theircausal relations without losing the quality of the results. Application: The CREAM links can be applied to human erroranalysis in any industry with minor modifications.