Patients’ common opinion and the results of studies conducted in Poland have shown that the level of satisfaction is higher among patients using the services of non-public health units. In the literature available the author has neither encountered any reports concerning the evaluation of this problem with respect to health providers in the Sub-Carpathian Region, nor an unequivocal reply to the question: what exerts the main effect on the level of perception of the quality of medical services among patients using public and non-public health care units. The objective of the study was the identification of the elements of the quality of ambulatory health services (within the scope of primary and specialist health care) which exert a decisive effect on the selection by patients of either public or non-public health unit. The study was conducted in 2007 by the method of stratified sampling (four groups of health units), and for each group, 10 units were selected at random from among those which had a contract signed with the Sub-Carpathian Regional Agency of the National Health Fund with its seat in Rzeszów, which have been active in the area of the Sub-Carpathian Region for at least one year. In each of the units selected by the above-mentioned method the surveyors selected the respondents according to an established key (instruction). A total number of 577 people were examined who used ambulatory primary and specialist health services provided by public and non-public health care units. The research technique was a direct standardized survey carried out based on uniform questionnaire forms designed by the author. Statistically significant relationships were identified with the use of the following statistical tools: chi-square test, Pearson’s correlation coefficient and one-way analysis of variance - ONEWAY ANOVA. Ultimately, the net effect of individual variables was assessed by means of logistic regression analysis. In order to make the analyses more clear, patients’ replies to the questions concerning the evaluation of individual elements of the quality of medical services were grouped in such a way that 5 synthetic scales were created with reliability confirmed by Cronbach alpha: 1. level of medical services; 2. level of conditions in the outpatient department and provision of services by the staff of the pa-tient registration room; 3. quality of medical information provided concerning the health problem of a patient; 4. possibility of obtaining the primary scope of diagnostic-therapeutic services in the health unit examined (basic laboratory tests, ECG, nursing procedures); 5. possibility of obtaining in the health unit examined of a expanded scope of diagnostic-therapeutic services (Rtg, USG, physiotherapy), advice by other specialists and referrals to these procedures. A statistically significant effect on the change of a chance of the selection by patients of either public or non-public unit was exerted by changes in only the last three synthetic scales. While selecting non-public health care units the patients were guided primarily by the possibility of obtaining from a physician complete information concerning their health problem. Patients who decided to receive treatment in a public health unit considered primarily the possibility of obtaining in this unit complex diagnostic-therapeutic services, both within the basic and expanded scope of services, as well as advice of other specialists and referrals for the above-mentioned services. The level of services provided by medical staff, level of conditions in the unit and provision of services by the staff in the registration room were not the variables which statistically significantly differentiated the preferences of patients. While choosing an individual health unit, patients most frequently considered the distance from the place of residence (62%), and the reputation (professionalism) of a physician (22.2%). Significantly more rarely they considered information pertaining to good services provision by the staff (8.1%), good lodging conditions (1.4%), and advertisements in the media or in the form of brochures (0.7%), which was also confirmed by other studies.