Fecal samples submitted for virus examination over July 1990 to June 1991 from children <3 years of age were examined by electron microscopy (EM), virus culture (VC), and enzyme immunoassay [EIA, group-reactive and adenovirus (Ad) 40/41 specific; Cambridge BioScience] to compare the detection rate of adenovirus from pediatric fecal specimens. Ad isolates of serotypes 1–7 grown in HEp-2 or primary rhesus monkey kidney cells were identified by neutralization. Graham 293 cell cultures were used only when specimens were found to be positive for Ad by EM, type-specific Ad40/41 EIA, and for isolates not identified by neutralization. Ads grown in 293 cells were identified by DNA restriction endonuclease analysis. Of the 1187 specimens examined, 105 (9%) were found to be positive for Ad. VC detected 93, while 12 additional positives were detected by EM or EIA. The relative sensitivity of VC, EIA, and EM for the 105 specimens was 89% (93), 45% (47), and 35% (37), respectively. Among the 105 positive specimens, enteric Ad, nonenteric Ad, and untypeable Ad were 28% (29), 65% (68), and 7% (8), respectively. Of 37 EM positives, 62% (23) were enteric Ad; 27% (10) were nonenteric including serotypes 2, 3, 4, 5, 12, and 31, with 4, 1, 1, 2, 1, and 1 isolates of each type positive, respectively; and 11% (4) were detectable only by EM. Five isolates were identified as variant of Ad 2(3), Ad 3(1) and Ad 31(1). Over a 1-year period, a single Ad41 variant strain was the most frequently detected enteric Ad in Winnipeg, Manitoba, Canada. For maximum detection rates of Ad viruses in pediatric fecal specimens, a combination of EM, VC, and EIA is required, but group-reactive EIA, or EM followed by Ad40/41-specific EIA of initial positives, are the most direct and efficient methods for enteric Ad detection.