Background. Antibiotic prescribing is more prevalent in children. Many factors influence this practice, including the burden of outpatient visits. We aimed to compare antibiotic prescribing for children by low prescribers (LP) and high prescribers (HP) in primary care. Methods. We analyzed pediatric prescriptions in primary care in Istanbul. Among the physicians randomly selected by systematic sampling, those generating ≥1 pediatric prescription/day (n=1218) were defined as LP or HP when they belonged to the lowest (n=305) or highest (n=304) quartile of prescribing, respectively. The antibiotic prescribing characteristics of these groups were compared. Results. We identified that 38.5% of the prescriptions written by physicians included antibiotics, significantly higher in HPs (38.8%) than in LPs (37.2%), (p=0.04). Among antibiotic-containing prescriptions, the mean number of drugs and boxes and the percentage of prescriptions containing injectable drugs/antibiotics were significantly higher in HPs compared to that in LPs. We detected that co-amoxiclav was the most frequently prescribed antibiotic in the LP and HP groups (61.1% and 48.3%, respectively). Stratification of antibiotics by their spectra showed that 11.2% were narrow, 79.8% were broad and 0.5% were ultra-broad-spectrum drugs. LPs were significantly more likely to prescribe broad-spectrum antibiotics (82.5%) than do HPs (78.9%,p<0.001). Conclusions. Antibiotic prescribing remains excessive in pediatric primary care, slightly more marked in HPs. While HPs also tend to prescribe a higher number of overall and injectable drugs/antibiotics, broad-spectrum anti-biotherapy seems to be more practiced by LPs surprisingly. Both physician groups appeared to prefer either narrow- or broad-spectrum drugs without paying enough attention to their pharmacodynamic properties.