OBJECTIVES: To assess the validity of cost calculation from incomplete data collection and point out its consequence on bias and on precision of study results. METHODS: Over a one-year period resource use data from 234 elderly patients with myocardial infarction were collected quarterly. Applying unit costs, costs for each quarter were estimated. To examine different strategies of incomplete data collection, two different methods (inter-/extrapolation) with three different time frames (omitting quarter two and/or three) were applied and compared with complete data collection. Sample size was recalculated in response to the variation of cost estimates due to incomplete data collection. RESULTS: Sum of cost estimation from complete and incomplete data collection in the case of omitting information of quarter two or three showed no significant difference. When the time sampling included only 50% of the full information (omitting quarter two and three) costs were significantly lower by 3.9% (extrapolation) and 4.6% (interpolation). Generally, an increase in the standard deviation by 1% leads to an increase in sample size by 2% in the case of a single outcome. Thus, based on observed increased standard deviation due to incomplete cost collection a larger sample size by about 3% would be required. This would be efficient, since more time is saved due to incomplete data collection than extra time is required for additional patients. CONCLUSIONS: In economic evaluation, cost data can be collected efficiently by reducing frequency of data collection. This can be achieved by data collection for shorter periods (incomplete data collection) or extending recall windows (complete data collection). When applying the method of incomplete data collection, sample size calculation has to be modified due to increased standard deviation. This approach is suitable to enable economic evaluation with lower costs to both study participants and investigators.