Abstract. A grouped data model for Weibull distribution is considered. Under mild con- ditions, the maximum likelihood estimators(MLE) are shown to be identifiable, strongly consistent, asymptotically normal, and satisfy the law of iterated logarithm.