This paper investigates pilot optimization in cell-free massive multiple-input multiple-output (CF mMIMO) systems under the \(\kappa -\mu \) shadowed fading model. This study aims to optimize the pilot allocation algorithm to maximize the number of supported user equipments (UEs) while meeting the stringent error probability (EP) and the latency requirements of massive ultra-reliable and low-latency communication (mURLLC) systems. The \(\kappa -\mu \) shadowed fading model, characterized by the parameters \(\kappa \) , \(\mu \) , and m, encapsulates both multipath and shadowing effects, thereby generalizing classical fading models such as Nakagami-m fading, Rayleigh fading, and Rician fading. This model’s flexibility enables more accurate characterization of channel properties across diverse wireless environments. We derive the signal to interference and noise ratio(SINR) under the \(\kappa -\mu \) shadowed fading model, employing least squares (LS) channel estimation and maximum ratio combining (MRC) detection. Next, we calculate the EP via finite blocklength(FBL) theory. Following this, we propose the Dynamic Pilot Optimization and Allocation (DPOA) algorithm, which iteratively adjusts the pilot length to satisfy the mURLLC performance requirements. The simulation results demonstrate that the DPOA algorithm, through efficient pilot reuse, satisfies mURLLC performance criteria while utilizing fewer pilots, thereby accommodating the same number of UEs across various shadowed fading scenarios.