In the last decade, multidrug-resistant tuberculosis (MDR-TB, defined as resistance to at least isoniazid and rifampicin) has become an epidemiological issue of first priority at the global level. Case management needs to be simplified and standardised, as in many countries MDR-TB cases cannot receive individualised attention from specialist physicians. However, before any decision can be made on standardisation, a careful analysis must first be made of the evidence and controversies behind the various published recommendations. Unfortunately, the controversies outweigh the evidence. The difficulties lie not only in the absence of controlled trials to validate specific recommendations, but also in the very different and even contradictory results found in the literature. It is therefore essential to analyse these discrepancies before developing rational, uniform recommendations. The analysis should encompass the most essential and controversial issues regarding the management of MDR-TB patients: 1) confirmation of diagnosis in a suspected MDR-TB patient, and determination of the value of drug susceptibility testing; 2) the number of anti-tuberculosis drugs required to treat MDR-TB; 3) the most rational use of effective drugs against tuberculosis; 4) the advisable length of parenteral drug administration or of the initial phase of treatment; 5) the contribution of surgery to the management of MDR-TB patients; and 6) the optimal regimen for treating MDR-TB: standardised vs. individualised regimens. The evidence and controversies regarding each of the above questions are analysed with the aim of facilitating decision making in the treatment of these complex patients.