This work considers the parameter and state estimation of a low-order model of anaerobic digestion of cattle dung. First, on the basis of a previously identified parameter set, partial parameter estimation with laboratory experimental data is performed. A modern numerical sensitivity analysis, to our knowledge not used for microbial biotechnology models so far, is a successful part of the estimation procedure. Second, the estimation results are validated with pilot-scale experimental data and can be considered satisfactory. Third, state estimation of the main AD variables is made via H-infinity filtering.