A field experiment was conducted for two consecutive years (2013 and 2014) during the dry periods under irrigation at Dabat District, Northwestern Ethiopia to select the best adaptable and high yielder varieties. Five garlic varieties (BishoftuNech, Kuriftu, Tseday 92, MM-98 and the local) were planted on randomized complete block design with three replications. Data on days to emergence, days to maturity, pseudo stem height, number of leaves per plant, bulb diameter, number of cloves per bulb, fresh bulb weight, bulb yield, and bulb dry matter content were collected and subjected to analysis of variance (ANOVA) using SAS computer software version 9.0. Mean separation test was done by list significant difference (LSD) test at 5% probability level. The results revealed that varieties were significantly differed in all parameters. BishoftuNech was early maturing variety (135 days) followed by Kuriftu (143 days). MM-98 was matured lately at 176 days after planting. Significantly highest pseudo stem height and number of leaves per plant were recorded from the local variety. Similarly, significantly highest yield of 16.56 t·ha^﹣1 was recorded from the local variety. MM-98 and the local variety were with the highest bulb dry matter content of 25.83 and 23.86%, respectively. This research proved that the improved varieties were not compared with the local cultivar before they get released. Therefore, the result of this research can be used as good information for the future garlic variety development program at national level. Considering the most desirable yield and yield component parameters, the local variety is recommended to the producers at Dabat District. Further research on collection, characterization and evaluation of the local cultivars should be conducted for national use.