Introduction: We aimed to identify factors affecting percutaneous nephrolithotomy (PNL) outcomes among patients with horseshoe kidney (HSK) and assess the predictive accuracy of the Clinical Research Office of the Endourological Society (CROES), Guy’s Stone Score (GSS), and S.T.O.N.E. scoring systems. Methods: Data from 98 patients with HSK who underwent PNL between November 2010 and January 2020 were evaluated. Patients were divided into the stone-free (SF) and non-SF groups and compared according to demographic data, stone and surgical characteristics, and stone scoring systems. Multivariate logistic regression analysis was performed to identify factors associated with SF status. Results: Among the included patients, 87 were male and 11 were female (mean age: 47.37 ± 14.42 years). The SF rate was 84.7% (83 patients). Group analysis identified GSS (p < 0.001), CROES score (p < 0.001), S.T.O.N.E. score (p = 0.014), stone burden (p = 0.045), and multiplicity (p < 0.001) as factors associated with SF status. Among our cohort, 10 patients developed complications. All scoring systems were significantly correlated with SF status (CROES: r = −0.442, p < 0.001; GSS: r = 0.442, p < 0.001; S.T.O.N.E.: r = 0.250, p = 0.013), while CROES score was identified as an independent factor associated with SF status (95% CI: 0.937–0.987; p = 0.003). Conclusions: PNL is an effective method for treating nephrolithiasis among patients with HSK. Moreover, stone-related factors, such as larger size, multiplicity, and complexity, were associated with procedural failure. Finally, the CROES nomogram was a better predictor of SF status compared with other scoring systems.