In China, maize (Zea mays L.) hybrids are routinely evaluated in multi-environment trials and stability of performance is considered important. The objectives of this investigation were to: 1) evaluate stability of performance of 13 Chinese hybrids tested across 10 locations in Yunan in 2002 and 2003 via GGEbiplot and yield-stability statistic (YS_i), and 2) identify non-representative and non-discriminating locations. Within years, hybrids and hybrids-by-locations (H x L) interaction were significant. The partitioning of H x L interaction into heterogeneity and residual interaction revealed that environmental index did not contribute appreciably to interactions. GGE biplot analysis ranked hybrids with above average yield across years for yield as: Hai He>LD10>YZ1>Tun004 and for stability of performance as: LD10, Hai He, Tun004, and YZ1. YS_i identified, among the top five hybrids, LD10, Hai He, and YZ1 as common between years. Both GGEbiplot and YS_i identified QC3, XHD892ck, and R313 as the least desirable hybrids. YS_i indicated ZZY6 and SB21-3 to be the most-unstable hybrids between years. The only hybrid with a non-significant stability variance (σ_i²), i.e., stable performance, across locations was Tun004 in 2002. All other hybrids showed a significant σ_i² in both years. The overall YS_i showed a relatively high correlation with GGE distance (r = -0.92**). Thus, in the absence of GGEbiplot software, YS_i may be reliably used for selecting hybrids with high yield and stable performance. GGE biplot analyses revealed that out of 20 possible year-location environments, Hai He had the highest yield in 7 and LD10 in 10 environments. GGEbiplot identified two locations in 2002 and one location in 2003 to be the least discriminating and three common locations in both 2002 and 2003 to be the least representative of all test locations. Redundant locations were also identified. These findings can help breeders/agronomists fine-tune the maize-testing program in question.