During the past 10 years 311 consecutive patients were admitted with 342 episodes of small bowel obstruction (SBO). There were 193 cases of partial small bowel obstruction (PSBO) and 149 cases of complete small bowel obstruction (CSBO) as determined by interpretation of the abdominal radiographs done on admission. The purpose of this review was to determine the reliability of the admission plain abdominal radiographs and subsequent upper gastrointestinal (UGI) contrast studies in predicting the need for operative intervention. The use of nasogastric tubes (NGT) versus nasointestinal (long) tubes (NIT) was correlated with the following outcome variables; length of hospital stay (LOS), timing of operative intervention, incidence of postoperative complications, and duration of postoperative ileus. Long tubes (NIT) were used in 64 episodes of PSBO and 81 episodes of CSBO, whereas nasogastric tubes (NGT) were used in 116 cases of PSBO and 68 cases of CSBO. Thirty-eight of 193 (19%) patients with PSBO required operation (20 of 116 with NGT and 18 of 64 with NIT), whereas 125 of 149 (84%) patients with CSBO required operation (60 of 68 with NGT and 65 of 81 with NIT). Need for operation was not correlated with whether or not long tubes passed beyond the pylorus; 50 passed versus 33 not passed in operative groups (p = 0.15). Twelve of 83 patients with NIT had operation within 24 hours versus 52 of 80 patients with NGT (p less than 0.001). In six of 64 patients who had surgery within 24 hours, complications developed versus in 39 of 99 patients operated on more than 24 hours after admission (p less than or equal to 0.001). In 29 of 83 patients treated with NIT, postoperative complications developed versus in 16 of 80 patients with NGT (p less than or equal to 0.04). The mean duration of postoperative ileus in patients with NIT was 7 days versus 4.1 days for NGT patients (p less than 0.001). The mean LOS was 12.2 days for NGT patients versus 21 days for patients with NIT (p less than 0.001). Barium UGI contrast studies were performed in 57 patients to establish the presence of obstruction. In 34 of 57 patients the UGI disclosed mechanical obstruction that required operative intervention. In the remaining 23 patients no obstruction was demonstrated, and all 23 patients recovered without operation. In conclusion, there is no inherent superiority of NIT versus NGT in the treatment of SBO.(ABSTRACT TRUNCATED AT 400 WORDS)