OBJECTIVES: There is an increasing need to inform healthcare resource allocation in an efficient way. A legitimate question for LA policy makers is to what extent economic evaluations of healthcare technologies (EEH) conducted both in industrialized economies and in LA are generalizable both to LA (trans-regional) and to other LA countries (intra-regional). The objective was to evaluate both trans-regional and intra-regional generalizability of EEHs; as well as to identify the barriers to the geographic transferability of the results. METHODS: Systematic review (NHS EED, OHE HEED, LILACS and NEVALAT) to identify trial and model-based EEHs (1980-2004). We partial (cost-consequences) and full (CEA, CUA and CBA) EEHs which involved at least one LA country. Data extraction was done by three independent reviewers. RESULTS: 73 EEHs were included (38% RCTs, 29% models, 19% non-randomized studies and 14% mixed trial-modeling approach). Reporting quality of the 49 trial-based EEHs was heterogeneous; in 31% sample characteristics were not reported, and the follow-up period was not reported. Health benefit was based on the full original study sample in 88%, however patient level resource use was used only in 67%. Resource use and unit costs were usually not reported separately; and costing details were generally insufficient to perform local present value analysis. Most multinational trials reported aggregate clinical data, resource use, and costs in a way that limited the assessment of generalizability to LA. Exploration of geographical variation of results was almost uniformly absent. A similar trend was observed among the 32 model based EEHs. Most of the few CUAs used preference weights from non-LA settings. CONCLUSION: The analysis and reporting of patient level clinical and economic data do not closely adhere to existing international guidelines. Although model-based EEHs are somewhat more rigorous; the generalizability of findings is rarely explored. Both trans-regional and intra-regional generalizability is usually weak in current EEHs.