Onion varieties were evaluated for adaptability to the climatic conditions in Botswana. Six varieties of onion were planted under a complete randomized block design experiment with four replications at farmers' field in Botswana (Southern Africa). The quantitative data on yield of onion varieties and, farmers perception on the varieties were collected. All the varieties were harvested at their maturity (80% tops down) and then graded and weighed into six categories of bulbs: small weight, double weight, thin weight, bolters weight, marketable weight and total weight. Inferential and descriptive statistical methods were used for data analysis where by two ways analysis of variance and ranking scales were applied. It was concluded that all the six varieties namely; Arad, Galil, Hanna, Shahar, Texas Grano 502 P.R.R. and Red Creole contributed differently to the total yield, marketable yield and percentage marketable yield. Onion variety Hanna followed by Shahar produced the highest marketable yield whereas Galil variety gave the lowest marketable yield. Despite the fact that Texas Grano variety produced the highest total yield and has been a common and popular onion variety in Botswana, it was recommended that farmers should not grown Texas Grano variety as it produced the lowest percentage (60%) of marketable yield. Based on the yield performance and the farmer's perception, Hanna followed by Shahar varieties were found to be the best performing onion variety and therefore, government should promote these onion varieties for cultivation in Botswana. Government can import to ensure the availability of Hanna and Shahar varieties so that farmers can be motivated for adoption of these varieties for commercial production. The effects of different planting dates on the yield of onion varieties can be considered for future research.