PURPOSE OF THE STUDY The purpose of the study is a retrospective comparison of results of the two-stage revision total hip arthroplasty using a non-articulating and an articulating spacer to treat periprosthetic joint infection (PJI). Two basic hypotheses are evaluated: (1) the clinical outcomes of the patients treated with "hand made" articulating cement spacer are better than in non-articulating patient's group in two-stage revision for PJI of the total hip arthroplasty and (2) PJI recurrence is higher in the group of patients treated with an articulating spacer group. MATERIAL AND METHODS The evaluated group consists of a total of 57 patients (23 women, 34 men) with the mean age of 61.2 years. Group A of 39 patients were treated by two-stage revision using the "hand-made" articulating cement spacer and Group B of 18 patients were treated using the non-articulating spacer. Both the groups were evaluated retrospectively in the reference period: preoperatively and two years after the surgery using the Harris Hip Score (HHS) clinical assessment. The revision surgery for acute and chronic complications of treatment, length of hospitalization, and the PJI recurrence were evaluated for both the groups. RESULTS The resulting HHS clinical reviews were pre-operatively 43.59 points in both the groups with postoperative improvement up to 81.74 points. The mean preoperative HHS scores were 41.67 points (Group A) and 47.77 points (Group B) and two years after the surgery they were 83.43 points (Group A) and 78.08 points (Group B) (two-tailed t-test, p-value = 0.042). In Group A a total of seven revisions were performed in the interval between the two-stage revision (4x recurrent dislocation, 2x persistent infection, 1x spacer fracture). In Group B one patient was revised for persistent infection. In the two-year period after the operation, a relapse of PJI was recorded in 5 patients in Group A (12.8%) and in 1 patient in Group B (5.6%) (Chi-square test, p-value = 0.41). The average time of hospitalization was 51.58 days, whereby 49.72 days and 55.61 days on average for Group A and B respectively (p-value = 0.53). DISCUSSION According to recent studies, the advantage of motion preservation in articulating cement spacers can be complicated by recurrent dislocations, implant migration, periprosthetic fractures or recurrent joint replacement infections, which can further prolong the treatment and worsen the final clinical results. An alternative treatment option is the application of a nonarticulating spacer maintaining the advantage of local administration of antibiotics and reducing the dead space formed by the infected implant removal. Discussed is mainly the choice of the method in case of muscle disorder or presence of segmental bone defects. CONCLUSIONS The results demonstrate the better clinical outcomes and the higher revision rate of patients with an articulating cement spacer in two stage revision. We didn't find any differences between the risk of PJI recurrence in both groups. Key words:periprosthetic infection, total hip replacement, cement spacer, two stage revision, articulating spacer, nonarticulating spacer.