OBJECTIVES: The NICE methods for technology appraisal require systematic literature reviews across clinical, cost-effectiveness, cost and resource use and health state utility values. These searches are a key component of a manufacturer submission. Whilst differences in population terms are expected across appraisals, this study reviewed search strategies submitted in three recent NICE appraisals to assess whether search terms used for methodological filters were consistent. The choice of databases searched were also reviewed. METHODS: NICE technology appraisals were reviewed in reverse chronological order. Manufacturer submissions were searched to identify those that reported the search terms used. In order to keep the search manageable, a target of three papers was set in the first instance. Search terms for the following methodological filters were extracted: quality of life, cost and resource use, and cost-effectiveness. Results were compared qualitatively to assess the consistencies across search terms. Also extracted were the databases searched, presentation of search protocols and any Evidence Review Group (ERG) comments. RESULTS: Search terms used for the methodological filters varied across submissions and similarly the databases searched differed. This was noted for all searches, with the comprehensiveness of searches differing across the three appraisals. In consideration of the presentation of search protocols, there was further variation resulting in lack of transparency. A common approach was to run a single search strategy for the cost-effectiveness and cost and resource use reviews. CONCLUSIONS: From this initial review there appears to be a lack of consistency in search terms included and databases searched in NICE technology appraisals. Whilst generic search filters are available there appears to be a need to drive awareness and application of these search terms. The implication for technology appraisals is decision making based on non-optimized evidence bases.