Even though nicotinic acid (niacin) appears to have beneficial effects on human lipid profiles, niacin-induced cutaneous vasodilatation called flushing limits its remedy to patient. GPR109A is activated by niacin and mediates the anti-lipolytic effects. Based on the hypothesis that β-arrestin signaling mediates niacin-induced flushing, but not its anti-lipolytic effect, we tried to find GPR109A agonists which selectively elicit Gi-protein-biased signaling devoid of β-arrestin internalization using a β-lactamase assay. We identified a 4-(phenyl)thio-1H-pyrazole as a novel scaffold for GPR109A agonist in a high throughput screen, which has no carboxylic acid moiety known to be important for binding. While 1-nicotinoyl derivatives (5a-g, 6a-e) induced β-arrestin recruitment, 1-(pyrazin-2-oyl) derivatives were found to play as G-protein-biased agonists without GPR109A receptor internalization. The activity of compound 5a (EC50 = 45 nM) was similar to niacin (EC50 = 52 nM) and MK-6892 (EC50 = 74 nM) on calcium mobilization assay, but its activity at 10 μM on β-arrestin recruitment were around two and five times weaker than niacin and MK-6892, respectively. The development of G-protein biased GPR109A ligands over β-arrestin pathway is attainable and might be important in differentiation of pharmacological efficacy.