PURPOSE OF THE STUDY Supracondylar humerus fracture (SCF) with dislocation is indicated for closed reduction and osteosynthesis. The method achieving the best stability is CRCPP (closed reduction and crossed percutaneous pinning), even though there is a risk of iatrogenic ulnar nerve injury. The CRLPP (closed reduction and lateral percutaneous pinning) method eliminates this risk at the cost of less stable osteosynthesis. The purpose of this study is to compare the SCF stabilisation by CRLPP with the stabilisation by CRCPP in rotationally stable fractures and to identify the risk of iatrogenic ulnar nerve injury, or the failure of osteosynthesis with recurrent dislocation of fragments. MATERIAL AND METHODS The prospective group of the patients with SCF type 1/2 (classification according to Havránek) treated in the period 2016-2018, in whom the method of osteosynthesis (number of implants, method of their insertion), resulting condition and complications (nerve injury, failure of osteosynthesis) were evaluated. In the second half of the study, in CRLPP one of the implants was inserted "quadricortically", i.e. through the olecranon fossa of the humerus (hereinafter referred to as fossa), while until then both the implants had been inserted through the radial column outside fossa. RESULTS In the period 2016-2018, 791 patients with SCF were treated at our department. In 225 cases (28.5%) the patients sustained the type 1/2 fracture and in all the cases closed reduction and percutaneous osteosynthesis were performed, namely CRCPP in 185 cases (82.2%) and CRLPP in the remaining 40 cases (17.8%). Signs of ulnar nerve injury after osteosynthesis were observed in 35 patients (15.6% of SCF 1/2), always after the use of at least one ulnar implant (18.9% of CRCPP). A failure of osteosynthesis occurred in 2 cases (0.9% of SCF 1/2), always when only lateral implants were used (5% of CRLPP). DISCUSSION In both the patients in our study in whom after CRLPP a failure of osteosynthesis with rotational dislocation occurred, the original CRLPP was performed by inserting both the implants through a single column outside fossa. Both the patients were indicated for revision reduction and osteosynthesis was subsequently performed through CRCPP. The patients healed with no further complications. In the group of patients with an ulnar nerve injury, the original condition was fully restored, after 3.6 months (range of 1-10, median 3) on average. The results of our study show the need to guide the implants inserted through the radial column divergently so that they are at the fracture line level as far apart as possible (with adequate fixation of fragments). One of the implants is inserted through fossa, i.e. quadricortically. Based on our experience, the compliance with these principles alone shall ensure adequate rotational stability of SCF of type 1/2. In CRLPP, after the insertion of implants the stability is tested under the Xray image intensifier intraoperatively so that a medial implant can be added in case of unstable osteosynthesis. CONCLUSIONS Based on the results of our study we recommend to stabilise the rotationally stable SCF (type 1/2 according to Havránek) only from the radial column (and thus eliminate the risk of iatrogenic ulnar nerve injury), provided the fracture characteristics allows so. Nonetheless, the CRLPP has its own specific rules for implant entry which have to be adhered to. Key words: supracondylar fracture of the humerus, paediatric fractures, closed reduction, percutaneous pinning, lateral percutaneous pinning, iatrogenic ulnar nerve injury, osteosynthesis failure.