The objective of this study was to investigate the association between antimicrobial agent use and antimicrobial resistance in Escherichia coli isolated from healthy pigs using data from 2004 to 2007 in the Japanese Veterinary Antimicrobial Resistance Monitoring System (JVARM). Fecal E. coli isolates from 250 pigs (one isolate each from a pig per farm) were examined for antimicrobial resistance. Information on the use of antimicrobials within preceding 6 months and types of farms recorded in JVARM was collected and statistically analyzed against the resistance patterns. In the univariate analysis, associations between both therapeutic and feed additive use of antimicrobials, and resistance to dihydrostreptomycin, gentamicin, kanamycin, ampicillin, cefazolin, ceftiofur, oxytetracycline, chloramphenicol, trimethoprim, nalidixic acid, enrofloxacin, colistin, and bicozamycin, and husbandry factors were investigated. In multivariable analysis, generalized estimating equations were used to control geographical intraclass correlation. Confounding for structurally unrelated associations was tested using generalized linear models. The results suggested direct and cross selections in the associations between use of aminoglycosides in reproduction farms and resistance to kanamycin, use of tetracyclines in larger farms and resistance to oxytetracycline, use of beta-lactams and resistance to ampicillin, use of phenicols and resistance to chloramphenicol, and use of fluoroquinolones and resistance to nalidixic acid and enrofloxacin. Coselection was suggested in the use of tetracyclines and chloramphenicol resistance. The associations between use of beta-lactams and dihydrostreptomycin resistance, use of macrolides and ampicillin and oxytetracycline resistance, and use of colistin and kanamycin resistance were significant, but were confounded by the simultaneous use of homologous antimicrobials.