The combination of gemcitabine plus paclitaxel has been proposed as an alternative to the platinum-based combinations for treatment of advanced non-small-cell lung cancer (NSCLC). However, conflicting results have been reported. This meta-analysis was performed to compare the activity, efficacy, and toxicity of gemcitabine plus paclitaxel versus carboplatin plus either gemcitabine or paclitaxel in patients with untreated advanced NSCLC. Randomized phase II and phase III clinical trials comparing gemcitabine plus paclitaxel with carboplatin plus gemcitabine or paclitaxel were collected from electronic databases (Medline, EMBASE, and the Cochrane Central Register of Controlled Trials), relevant reference lists, and abstract books. The published languages and years were not limited. Pooled odds ratios (ORs) were calculated for the 1-year survival rate (1-year SR), the overall response rate (ORR), and grade 3 and grade 4 toxicities. Four randomized controlled trials (2186 patients) were identified from 2051 reports. They were all published as full-text articles. No significant heterogeneity was detected in these studies. A significant difference in ORR favoring gemcitabine plus paclitaxel over carboplatin-based doublets was observed [OR = 1.20; 95% confidence interval (95% CI) = 1.02-1.42; P = 0.03], whereas the trend toward an improved 1-year SR was not significant (OR = 1.07; 95% CI = 0.91-1.26; P = 0.41). An increased risk of grade 3-4 toxicities for patients receiving carboplatin-based chemotherapy was statistically demonstrated. The gemcitabine plus paclitaxel combination showed an improved ORR and a better toxicity profile but a similar 1-year SR compared to carboplatin-based doublets. For nonplatinum-based chemotherapy, gemcitabine plus paclitaxel is a useful alternative.