This paper presents a methodology for incorporating non-uniform measurements in state estimation for Linear Parameter Varying (LPV) systems. The methodology is based on a modified receding horizon state estimator. The proposed approach allows to explicitly estimate the state, reducing the estimation error, by including measurements that are not necessary available at every sampled time. A discussion of the existence of solutions and implementation issues are also presented. Finally, the methodology is illustrated in a numerical simulation for state estimation in a production process of δ-endotoxins of Bacillus thuringiensis (Bt).