A class of tests is considered for the -samples homogeneity hypothesis by interval censored failure time data. I.e., for each subject a random interval is observed to which its failure time belongs. The test statistics is
where is the number of subjects in the union of all samples, is the vector of indicators of the sample (its -th element equals 1 iff the -th subject belongs to the -th sample and is 0 otherwise), is a nonparametric estimator of the survival function under the null hypothesis (homogeneity), is a fixed function. (E.g., for this is the score statistics). The asymptotic normality of under is demonstrated. Simulation results and real breast cancer data application are considered.