objective: correlate the evolution of the resistance of staphylococcus aureus collected from healthcare workers with the local consumption of antibiotics. materian and methods: open prospective research.study site. general reference hospital with 200 beds in a 700,000 inhabitant region, in ponta grossa, paraná, brazil. results: two collections (samples) of staphylococcus aureus isolates were obtained from healthcare-workers during an approximate four-year interval. samples 1 (n= 200) and 2 (n= 270) had this bacterium in 63 (32%) and 90 (33%) of the patients, respectively. at the same time, the annual consumption of antibiotics in ddd/1,000 patient-days was determined. the variation of resistance was significantly smaller (m.s.d.=12.11) for gentamycin (p<0.01) and (m.s.d.=9.22) for tobramycin (p<0.05). the correlation between variation in resistance and antibiotic consumption was not significant. workers studied in the two samples showed a significant (p<0.01) frequency (c2=10.44) for persistent nasal carriage and for non carriage. methicillin resistant staphylococcus aureus was found in 12 (6%) patients of sample 1 and 11 patients (4%) of sample 2. conclusion: stability of resistance allows us to maintain therapeutic outlines. the variation in bacterial resistance in the twice-sampled population (n=105) indicated the selection pressure of the hospital environment. the resistance that was found is representative of the hospital microbiota; this relationship represents a biological model, based on the healthcare-workers' interaction with colonizing bacteria and nosocomial infections. new studies could improve this model for other bacteria, to determine the tendency for resistance and help guide the antibiotic use.