Semiquantitative (Maki) and quantitative (Brun- Buisson) culture techniques were employed in the diagnosis of catheter-related bloodstream infections (CRBSI) in patients who have a short-term central venous catheter (inserted for 30 days). The diagnosis of CRBSI was based on the results of semiquantitative and quantitative culture of material from the removed catheters. Catheter tips (118) from 100 patients were evaluated by both methods. Semiquantitative analysis revealed 34 catheters (28.8%) colonized by ≥15 colonyforming units (cfu), while quantitative cultures (34 catheters, 28.8%) showed the growth of ≥103 cfu/mL. Bacteremia was confirmed in four patients by isolating microorganisms of identical species from both catheters and blood samples. Using the semiquantitative culture technique on short-term central venous catheter tips, we have shown that with a cut-off level of ≥15 cfu, the technique had 100.0% sensitivity, specificity of 68.4%, 25.0% positive predictive value (PPV) and 100.0% negative predictive value (NPV), efficiency of 71.4% and a prevalence of 9.5%. The quantitative method, with a cut-off limit of ≥103 cfu/mL, gave identical values: the sensitivity was 100.0%, specificity 68.4%, positive predictive value (PPV) 25.0%, negative predictive value (NPV) 100.0%, efficiency 71.4% and prevalence 9.5%. We concluded that the semiquantitative and quantitative culture methods, evaluated in parallel, for the first time in Brazil, have similar sensitivity and specificity. Keywords: central venous catheter; semi-quantitative culture; quantitative culture; catheter-related bacteremia.Departamento de Análises Clínicas, Faculdade de Ciências Farmacêuticas, Universidade Estadual Paulista, UNESP, Araraquara, SP, Brasil.Departamento de Ciências de Computação e Estatística, Instituo de Biociências, Letras e Ciências Exatas, Universidade Estadual Paulista, UNESP, São Jose do Rio Preto, SP, BrasilDepartamento de Análises Clínicas, Faculdade de Ciências Farmacêuticas, Universidade Estadual Paulista, UNESP, Araraquara, SP, Brasil.Departamento de Ciências de Computação e Estatística, Instituo de Biociências, Letras e Ciências Exatas, Universidade Estadual Paulista, UNESP, São Jose do Rio Preto, SP, Brasil