The aim of this study vvas to separate empty seeds and determine the effects of seed classification on germination of Crimean Juniper (Juniperııs excelsa Bieb.). In order to succeed in this aim, seeds that vvere naturally found in Aliköy of İsparta region vvere used. A pretreatment combination vvas applied for separating the empty seeds for 41 days. After the empty seeds are separated, the seeds vvere divided into 3 classes of size (I:<2.5 mm, II: 2.5 mm-3.4 mm, III: >3.4 mm). Follovving the classification procedure, seeds from 3 sub-lots vvere submerged again in saccharine Solutions in different concentrations to reach the highest level of viability. After reaching the highest viability, under the open-field conditions seeds from 3 sub-lots vvere sovvn in vvooden cases vvith a size of 60 cm x 80 cm for germination tests. “Completely randomised design” vvas used vvith three replications after applying 2 pretreatments (1: hot stratification at 20 "C for 30 days, 2: hot stratification at 20 °C for 30 days + hot stratification at 10 "C for 30 days). Germination data on seed classes and pretreatments vvere analyzed using SPSS statistical softvvare. Analysis of variance and Duncan test shovved that the seed pretreatments did not affect germination. Hovvever, there vvere significant differences among 3 seed classes in terms of germination percentage. The highest germination percentage (66 %) vvas achieved from the seeds of the first class vvhich vvere sovvn after applying hot stratification at 20 °C for 30 days and hot stratification at 10 "C for 30 days. In order to - increase the viability of the Crimean Juniper seeds, it is necessary to classify them according to their size and re-float them in saccharin Solutions vvith suitable concentrations.