Global broiler chicken pen trial reports (1993-2003) were analyzed statistically to determine effects of Saccharomyces cerevisiaevar. boulardii yeast outer cell wall mannan oligosaccharide (MOS; Bio-Mos®, Alltech, Nicholasville, Kentucky USA) supplemented diets versus negative control (nCON) or antibiotic-supplemented positive control (pCON) diets. Criteria for selecting studies were: 1) pen trial, 2) written report, 3) MOS fed for entire study period, 4) negative and/or positive control, 5) antibiotic stated (for positive control), 6) replication and 7) age, body weight (BWT) and feed conversion ratio (FCR) given. Mortality (MORT) was used when reported. Results were averaged "by treatments" (all comparisons) and "by trials" (comparisons averaged by trial before analysis) using Paired T-test to compare nCON and pCON means with corresponding MOS means. Slightly different answers but similar patterns emerged by these methods. Considering results averaged by trials, MOS diets gave the following relative improvements compared to nCON diets: BWT, +1.61%; FCR, -1.99% and MORT, -21.4 (all significant at P = 0.020). Relative improvements using MOS feeds compared to the pCON diets were: BWT, -0.36% and FCR, -0.11% (P = 0.473; nonsignificant differences). The MOS diets significantly (P = 0.008) lowered mortality (-18.10%) relative to pCON diets, indicating a strong beneficial effect. The MOS diets produced BWT and FCR comparable to those of pCON diets but significantly lowered MORT compared to antibiotic diets. Currently recommended optimal MOS levels of addition for broiler chicken feeds are: 0.2%, 0 to 7 days; 0.1%, 7 to 21 days and 0.05%, 21 to 42 days (or market).