PURPOSE OF THE STUDY Dislocation is one of the most common early complications of total hip arthroplasty (THA). In this manuscript, 20 years of experience with the management of this complication are presented, particularly in relation to the type of instability. MATERIAL AND METHODS In the period between January 1999 and December 2020, at least one dislocation occurred in 157 of 8 286 (1.9%) THA patients, of which 117 dislocations (1.6%) in primary and 40 (3.4%) in revision THAs. Almost all patients were operated on from the anterolateral approach during the follow-up period. The type of dislocation was evaluated using the modified Dorr classification. In the first dislocations, conservative approach was usually opted for, except for cases with a clear malposition, irreducible or unstable hips after the reduction. The minimum follow-up period was 18 months (18-240). The success rate of the chosen treatment approach was assessed by means of standard statistical methods. RESULTS The total dislocation rate in the follow-up period was 1.6% for primary THAs and 3.4% for revision THAs. The dislocation rate was slightly higher between 1999 and 2009 compared to the following decade (2.1% versus 1.3% for primary THAs; p=0.009). The most common type of dislocation was the positional dislocation (62%), followed by dislocations due to a combination of causes (17%) and component malposition (11%). Treatment of dislocation was successful in a total of 130 patients (130/157; 83%). Even though a stable hip was achieved in 21 patients (13%), the functional outcome was unsatisfactory, and in 6 patients (4%) we failed to achieve a stable hip. In the positional type of dislocation, the success rate of closed reduction following the first-time dislocation was 86.4% and a similar success rate was reported for reoperations in the first-time dislocations due to the malpositioned components (85.7%). In the second-time dislocation, the surgical therapy was significantly more reliable compared to closed reduction regardless of the type of dislocation (78.6% versus 46%). The treatment of dislocations following primary THAs showed comparable outcomes to those of the treatment of dislocations following revision THAs. Overall, the worst outcomes were achieved in patients with a combined type of dislocation. In total, the THA had to be removed in 11.5% of hips (18/157). The probability of final THA removal increased with the increasing order of dislocation. DISCUSSION In our group of patients, the dislocation rate in THA was comparable or lower than the published data. With the use of preventive measures, i.e. dual mobility cup or larger head diameters in high-risk patients, we managed to reduce the dislocation rate over time. The positional type of dislocation prevails in our group of patients just as in the previously published series, followed by instability from malposition of components. The modified Dorr classification is used to guide the treatment since it allows us not only to make good decision about the treatment modality but to some extent also to estimate the final outcome, particularly with respect to restoring a functional and stable hip. CONCLUSIONS The total dislocation rate was 1.6% for primary THAs and 3.4% for revision THAs. The first-time dislocation of the positional type shall be treated conservatively. Conversely, in the other types of dislocations and in recurrent dislocations, surgical treatment is more likely to achieve a good clinical outcome. The worst outcomes are to be expected in an instability due to combination of multiple causes, which leads to the removal of THA more often than in other types of dislocations. Also, the benefit of preventive measures in high-risk patients over time has been confirmed. Key words: total hip arthroplasty, dislocation, Dorr's classification, treatment strategy, outcomes, complications.