Hot cracking tests such as the Varestraint test are widely used for evaluating weld solidification and heat-affected zone cracking of austenitic stainless steels. Correlation between results of such simulative tests and actual practice is important because close assessment of weldability is required in welding cracking-sensitive materials. Several stainless steels including 316L, 316L(N) and a fully austenitic stainless steel D9 were evaluated using longitudinal and transverse Varestraint tests. These materials were also tested using the Finger Test, which simulates the strain situation during actual welding. Varestraint test results were quantified in terms of brittle temperature range (BTR) and crack length criteria. The results showed that cracking assessment using BTR was a strong function of WRC Cr-eq/Ni-eq ratio, while total crack length (TCL) showed high variability and scatter. Cracking assessment using TCL was improved by normalising with weld width, which reduced variability due to bead geometry. Results of the finger test showed excellent correlation with the Cr-eq/Ni-eq ratio and were more discriminatory for a highly susceptible fully austenitic material such as D9. The results of this study show how Varestraint tests can be used for weldability assessment with a high degree of confidence.