Introduction: Carpal collapse of wrist occurs in disorders like rheumatoid arthritis and Kienbock's disease. Three techniques have been described to measure carpal collapse. First, the carpal height ratio (CHR), measured by dividing carpal height by 3rd metacarpal length. Second, the revised carpal height ratio (RCH ratio), measured by dividing carpal height by length of capitate. Third, capitate radius distance (CR index), measured by shortest distance between distal edge of radius and the proximal edge of capitate. The index publications describe good reliability of all these but which method out of the three is best in terms of intra- and inter-observer variability is not known. The purpose of this study was to find out which method had the least inter- and intra-observer variability for determining carpal collapse. Materials and Methods: Fifty normal wrist postero-anterior radiographs were studied by three assessors who measured CHR, RCH ratio and CR index separately. The measurements were repeated after one month by all the three observers. The results were then statistically analysed. Results: The p-value was <0.001 in all the three assessors in CR index meaning that the intra-observer variability was least in CR index. For the inter-observer variability intra class coefficient of 0.9 indicated that the CR index has the least variability. Conclusion: CR index is the most reproducible method to measure carpal collapse. The method which provides accurate measurement of carpal collapse will allow better staging of carpal disorders.