A discovery of 'aspirin resistance' has prompted the search for fast and reliable methods for the monitoring of antiplatelet efficacy of acetylsalicylic acid (ASA). Our aims were: (1) to evaluate the in vitro model-based method for detecting a reduced platelets' sensitivity to ASA using a point-of-care platelet function analyser PFA-100; and (2) to propose a simple method of data analysis that might be successfully employed to discriminate between 'good' and 'poor' responders to aspirin. Whole blood platelets from healthy volunteers were incubated in vitro with 30 microg/ml ASA (the in vitro method under evaluation) or analysed following a 10-day intake of an average 150 mg ASA (Aspirin Protect) daily (the reference ex vivo method). According to polynomial regression analysis of the bimodally distributed data, the donors with lower ('ASA poor responders') or higher platelet sensitivity to ASA ('ASA good responders') were discriminated at 58.6% of platelet function inhibition. Despite the similar proportions of 'ASA poor responders' (44 versus 41% using the ex vivo and in vitro tests, respectively), 30% of discordant classifications point to a rather unsatisfactory convergence between both methods. Due to a considerable discordance between the in vitro and ex vivo tests of ASA efficacy performed with the use of the PFA-100 system, the former cannot be reliably and interchangeably used for the monitoring of aspirin therapy and the selection of an effective therapeutic ASA dose. A novel approach to data analysis of the distribution of platelet inhibition rates facilitates an evaluation of cut-off points required to discriminate between 'poor responders' and 'good responders' to aspirin.