INTRODUCTION: The aim of this study is to identify surgical, patient- and stone-related factors predictive of clinical success and complications after percutaneous nephrolithotomy (PCNL). MATERIALS AND METHODS: We prospectively studied 100 consecutive PCNL procedures. Univariate and multiple regression models were used in order to identify which variables could act as independent predictors of PCNL outcomes. Success was defined as complete absence of fragments in a non-contrast CT. The Clavien-modified grading system was used to classify the complications. RESULTS: Univariate analysis showed that patients rendered stone-free had a significantly lower stone burden, shorter operating times, single stones and non-struvite composed calculi. Patient age, nephrostomy tract dilation with high pressure balloon and a stone composition different to struvite behaved as significant protective factors for complications. Logistic regression models revealed that the main independent prognostic factor for success was stone surface (OR = 0.997 per mm2, p = 0.000), followed by multiple stones (OR = 0.203, p = 0.050). On the other hand, struvite composition (OR = 5.911, p = 0.028) was an independent predictor for the development of complications, whilst age (OR = 0.936, p = 0.012) and high pressure balloon dilation (OR = 0.041, p = 0.007) were rendered independent protective variables. CONCLUSIONS: Stone burden and multiple calculi in the kidney affect the immediate stone-free rate, whilst Amplatz dilation, struvite stones and young patients lead to a higher incidence of postoperative complications. This information can be very useful for patient counseling, regarding percutaneous kidney stone management.