Short tandem repeats are extremely abundant in the human genome and widely used in population and forensic genetics. The conventional approach to estimate mutation rates for STRs involves the detection of Mendelian incompatibilities and their frequency quantification, in family duos or trios. In this work, we aim to measure the impact this estimation methodology implies in the underestimation of mutation rates.We computed two independent experiences simulating one single-step mutation and one double step mutation in each pedigree. The pedigree with the highest number of compatibilities observed (hidden mutations) was mother-daughter duo, followed by the father-daughter duo, mother-son duo and, finally, mother-father-child trio, for the one and two-steps simulations. The estimation of X-STRs mutation rates from simple counting proved to be biased through systematic underestimation in an extent which depends on allele frequencies distribution of the marker, type of pedigree and number of repeat units involved.