Urinary tract infection (UTI) is one of the most common bacterial infections of childhood, with the incidence range between 1-1.5% in boys and 3-8% in girls. OBJECTIVE To estimate the value of urine analysis (bacteriology and biochemistry) for diagnostic of UTI in children. METHODS The method was to observe 700 urinary specimens collected from patients hospitalized in 4th Clinic of Pediatric Nephrology between October 2004 and February 2005, the urinalysis (UA) being performed in the Microbiology laboratory of "St. Maria" Emergency Children Hospital Iaşi. The sample collection was made from the midstream urine flow, direct microscopy and standard culture were performed immediately after collection. We applied a diagnostic strategy study in order to evaluate sensitivity and specificity of UA for prediction of UTI. The data was assessed according to the Receiver Operator Characteristic Curve. CONCLUSIONS We analysed the relation between leukocituria and positive urine culture and we find a sensitivity equal to 56%, with CI 95% = 0.56 +/- 0.07, and a specificity equal to 10%, with CI 95% = 0.10 +/- 0.02. The likelihood of UTI with negative UA result has been calculated as a negative likelihood ratio = 4.4. If we supplement a number of culture (2-3) we have obtain a sensitivity 97% +/- 0.02%, and the predictive negative value 97% (that mean the possibility to lose UTI in just 3% cases), and a specificity 35%, with a predictive positive value 61% (that mean a decrease of possibility to over-measure UTI). CONCLUSIONS The sensitivity of UA increase if we correlate more than leukocyturie and the significantly positive urine culture. We must to supplement a number of urine culture for increase the chance to obtain a positive diagnostic of UTI, and to exclude a possibility to over-measure UTI. Through increase a sensibility of method for diagnostic, decrease the possibility to loose and no treaty UTI with serious consequence.