Cannabis (Cannabis sativa) grown for flowers containing cannabinoids requires all female plants, which are susceptible to seed set from exposure to pollen. Created triploids demonstrated reduced seed production compared with diploids in field and greenhouse studies in which plants were challenged with pollen from males. In the field, seed production as a percent of floral biomass ranged from 6.7% to 18.0% for triploids and from 52.6% to 57.1% for diploids. The photoperiod-insensitive triploid genotype ‘Purple Star’ × ‘Wilhelmina’ had 98.5% fewer filled (containing a developed embryo) seeds than the photoperiod-insensitive diploid genotype ‘Tsunami’ × ‘Wilhelmina’. In the greenhouse, triploid ‘Wife’ had 99.5% fewer filled seeds than diploid ‘Wife’. Plant growth and flower production were similar with eight triploid and seven diploid genotypes evaluated over three greenhouse studies. There were a few superior triploid and diploid genotypes; however, their performance was more likely attributable to the parental cultivar combination than ploidy level. The optimal cross direction for producing triploid seed in large quantities is tetraploid × diploid because the diploid × tetraploid cross exhibits triploid block caused by endosperm paternal excess. Colchicine-induced tetraploid parent plants should be tested over a prolonged period to eliminate cryptic chimeral mixoploids or tetraploid plants should be derived from seed produced by crossing two colchicine-induced putative tetraploid plants to ensure that seeds from tetraploid × diploid crosses will be triploid. The latter approach is necessary for photoperiod-insensitive cultivars because a prolonged period of ploidy testing is not possible for these plants. These findings indicate that triploid plants have significantly reduced fertility and are a suitable alternative to diploids in situations in which pollen exposure is possible.